Fitting a Willis Immersion heater

tnegun

I'm planning on fitting one of these to use with an existing solar diverter, right now its connected to the original immersion in the tank but can only heat the top 1/3 - 1/2 the tank on bath. I was thinking of a longer immersion element but this looks like it has less chance of me damaging the tank! Instructions are for fitting are pretty vague but it looks like I take a 1/2" feed from the fill at the bottom of the tank and then return via a 1/2" at the top and let convection move the water/heat through the tank.

Has anyone fit one of these? There's a 1/2" drain on the filling feed at the bottom so I'm going to extend that to feed the Willis and then the feed from the tank is 3/4" so I'll source a 3/4"to 1/2" tee for the return from the Willis. Do I have the right idea?

John.G

Ensure its installed correcty, ie with the immersion element head pointing downwards, Google Willis immersion and it shows correct installation methods which are very important for the correct circulation rate,/temperature rise through the heater.

John.G

Google, external heater(willis) elemex.ie

tnegun

Thanks yup I saw a couple of posts re ensuring it's the correct orientation so that no air is trapped. I've ordered one now too and I think I've everything I need in my box of bits. Are penny valves ok to fit on the feed and return so that it can be isolated if ever required?

John.G

I would suggest fitting either gate valves of full bore ball valves, you can always throttle these valves in to give the correct temperature rise of ~ 40/50C through the heater.

The Continental Op

Just a question about valves. I can see they are a good idea for maintenance but are they a good idea when they could be closed off and the heater switched on?

tnegun

Good point I might just fit a valve on the supply side so the tank doesn't have to be completely drained to service.

John.G

It will be very interesting to see the variation in the temperature rise through the heater with the changing diverter heat inputs which will probably be well below the 3kw (or whatever) rating of the heating element.

tnegun

I'm using an Eddi so I think the max heater size is 3.6kw, my thinking right now is to leave the existing immersion as primary as this can heat the top 1/3 to 1/2 of the tank as it is and then when its stat cuts in the willis should do its thing and start a circulation in the tank adding more hot water at the top while drawing from the bottom.

Some of the diagrams show the willis connected to the feed coming off the top of the tank but I'm going to fit the return higher up on the expansion pipe. I think fitting it at the same height as the top of the tank would allow too much mixing when hot water is drawn off and the higher the return the lower the pressure from the willis reducing the amount of mixing?

John.G

Wonder what stops cold water bypassing the HW cylinder (via the willis heater) and mixing with the hot water being pushed out (like any hw cylinder) of the cylinder when a HW demand arises.

The Continental Op

Density? Cold water is heavier so won't move to the top of the pipes or tank.

The Willis heater is only an extension of the tank in that its fitted to the inlet at the bottom and the outlet at the top. Even if its off the water will stratify in the Willis and associated pipework just the same as it will in the tank.

John.G

No problem with understanding that with no HW demand but difficult to see how some water will not short circuit up through the willis when there is a hw demand as you effectively have two cylinders connected in parallel......maybe having the willis outlet teed into the expansion pipe higher up is enough?.

tnegun

I have that concern too which is why I'm fitting the return higher up on the expansion where it will have a foot or more less head so reduce the mixing and it will also be 1/2" vs 3/4" so that should help too. I was thinking partially closing the supply valve should help to restrict flow too. I think there will be some mixing regardless but these factors should keep it to a minimum does that make sense?

John.G

You can't throttle too much as the circulating force (head) due to convection is tiny and if you are dumping say 3kw into the heater then you must have a circulating flow of 1LPM to avoid the heater thermostat cutting out (and In). I havn't seen it (short circuiting) mentioned as a problem so install it and see what happens.

tnegun

Thanks for all the comments and guidance this far. The heater has arrived and I started figuring out the pipe work last night and have some obstacles to get around. I don't have a benders so am going to end up with a short rise from the top of the heater an elbow then 300mm horizontal a second elbow then a 300mm rise before the last elbow and 100mm horizontal to join the expansion. I think I might have problems with circulation/convection if installed like that?

John.G

I would do it in 3/4", in the unlikely event then that the circ flow is too high (low heater dT) can be catered for by throttling. Also try (not easy with elbows) to get get the horizontal sections at some, even small angle from the horizontal.

tnegun

Does it matter that the heater itself has 1/2" so I would be stepping down on the feed and back up again on the return?

John.G

I don't think so, give it a go, its only a bit of piping.

The Continental Op

Actually thats one thing that surprises me about the Willis Heater, I would have expected it to be a larger diameter 22mm pipes (- its from UK) to increase the flow.

tnegun

Just to report back I fit the heater and if anything it seems to be working too well as the water is getting dangerously hot. The stat is set to 60c but I think as the cold water is pulled in at the bottom the top is getting to 60c+ and being fed into the cylinder but the stat isn't reaching 60c. Flow through the Willis seems to be really good too I can feel the water circulating in the pipework if its going full tilt. Typically its now awkward to get to the stat but I'm thinking of setting it to 50c and see what impact that has.

cruizer101

That might actually imply there isn't enough flow i.e. water enters at bottom is getting heated but because there is low flow rate it spends a long time in the unit getting heated so gets very hot.

With a faster flow rate the water wouldn't spend as long being heated so wouldn't get so hot and the amount of water heated would increase.

If the thermostat is at the base it would make sense that it doesn't cut the immersion out. And the instructions would agree with that as they say to have a timer to stop it, e.g 15 minute timer.

The Continental Op

..

The Continental Op

https://www.boards.ie/discussion/comment/119046876#Comment_119046876

Don't think that will work? I would have expected the heating element will work exactly as you have described until the cold water being pulled in at the bottom is up to the temperature of the thermostat. You are in effect pumping X amount of heat into the tank which is stratifying as it should. The water below the stratified hot water may not get to 50C until the top of the tank is too hot. If the thermostat in the Willis is say a foot higher than the bottom the tank you have potentially to heat the water above the bottom foot of the tank before the thermostat becomes any use at all.

As I have said I've never used one (great to hear the feedback on its use!!! - there is so little online) but have always considered the thermostat to be more of last a ditch safety measure to stop the thing overheating than anything else. As suggested a timer seems to be the way to go.

John.G

https://www.boards.ie/discussion/comment/119046876#Comment_119046876

Is the hot water from the tap(s) too hot?? and is this by feel or measurement?.

Also did you use 1/2" or 3/4" piping on the outlet?.

tnegun

https://www.boards.ie/discussion/comment/119047000#Comment_119047000

I think you're right it has to be hoping to get a good flow through it.

https://www.boards.ie/discussion/comment/119047934#Comment_119047934

1/2" on the feed from the tank drain and then 300mm 1/2" on the return to get through a narrow spot before stepping up to 3/4" for the last 400mm and joining the 3/4" expansion pipe. I had another look tonight and if I changed routing slightly I could replace the 300mm 1/2" with 3/4" on the return if you think it would make a difference?

John.G

It could make that difference but the horizontal run(s) may be having a big effect if the circulation is too slow.

Is the cylinder actually heating up?? and as I asked above is the water overly hot coming from the tap(s)? because if the water is = or greater than 60C IN the cylinder then I can't see why the stat isn't cutting out.

The Continental Op

I'm looking at it this way. The water is getting really hot in the Willis then moving to the top of the HWC. The controlling factor for how hot the water gets is how fast the water moves across the heater. The faster it goes the less heat will end up in each unit of water. If increasing the pipe sizes will improve the speed of flow then the water in the tank won't be as hot.

It seems obvious from your information so far is that the Willis does exactly what it says on the tin https://heatingpartswarehouse.com/product/willis-external-immersion-heater/

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

The only simple solution I can think of is reducing the power of the heating element and try replacing the 3kW element with a 2kW one (rare but I did find one that was very expensive online so they must exist). However it might be that putting in a lower power element will just slow the heating down along with slowing the flow so the end result could be the same just slower?

Other solutions are to use a hot water temperature regulator to premix cold with your new found very hot water or to pump water through the Willis if a suitable pump could be found. At a guess you'd want a pump that would move the water only a little faster than its moving already from the quote you could say currently 4.5l of hot water is moving into the tank every 5 minutes so you want a pump that would force that up to say 6l every 5 minutes but a standard central heating pump would be far to fast.

+ As stated above the physical plumbing may have some room for improving the flow.

The Continental Op

https://www.boards.ie/discussion/comment/119048611#Comment_119048611

I wonder if it would be worth the OP running the Willis for exactly 5 minutes then switching it off (from a cold tank) and then measuring how much hot water he gets out of the nearest tap. If its a lot less than about 5 liters (4.5 from my quote) then we can assume the circulation is slower than it should be but if 5 minutes heating produces around 5 liters of hot water then the Willis heater is working as it should be.

tnegun

https://www.boards.ie/discussion/comment/119048611#Comment_119048611

Sorry yes it's too hot to hold your hand under, the first rise before a 90 turn is very short to the first horizontal so I suspect that doesn't help to get the circulation going. I'll see what I can do to replace the 1/2" in the meantime.

https://www.boards.ie/discussion/comment/119048622#Comment_119048622

Yes I agree the water is moving too slowly so very hot water sits on top of the tank I'll increase the size of the piping from the return and try get some angle on the horizontals and if it comes to it I'll look at adding a destratification pump.

https://www.boards.ie/discussion/comment/119048705#Comment_119048705

I can try boost for 5 mins, I'm using an Eddi solar diverter so I can't time the heater on and off for 10 minutes its on as soon as there is excess solar to go to the grid.

John.G

Yes, that's one way of checking it but as presumably the power source is from the diverter the supplied power will vary depending on solar panel output and in house use, the heating element could be connected up temporarily to the mains, the calculation is quite easy.

Circulation, LPM = (KWX860)/60/dT so if the cold water is 15C and the outlet is 60C then the dT is 45C so LPM = 3X860/60/45 = 0.96(LPM) so after 5 mins and IF the stat hasn't tripped the power then around 4.5/5 litres HW available. Ideally, to avoid the stat cutting in/out the dT should be ~ 35 to 40C, (flowrate 1.1 to 1.2LPM), the stat then won't cut the power until a full cylinder is hot and the return is say 20/25C.

John.G

https://www.boards.ie/discussion/comment/119048785#Comment_119048785

Increasing the pipework size/orientation is your best bet but just to ensure the stat is actually working I would reduce the setting to maybe 45C, obviously not a long term answer as the trick is to get the optimum dT until the stat is satisfied with a full (or timed) tank,