Willis Heater in an unvented system

[Siiimon]

At the suggestion of Alphonsox, I've been reading the descriptions of UFH systems heated by a willis heater. I'm seriously considering using one in our barn conversion in place of the intended wood burner. However, willis heaters appear to be only for vented systems and I'd rather have unvented for ease of install. How have people dealt with this?

[ProDave]

Who says they are only for vented systems?

[Siiimon]

Well that's what it says on the side of a popular make

[Guest Alphonsox]

I don’t think this is an issue. This is effectively a pressure restriction from the Willis point of view. A vented DHW system will run with the pressure from the cold water tank in the loft - say 1.5 bar max. Your UFH is unlikely to be running at anything over 1bar.

[Guest Alphonsox]

And welcome to the forum BTW.

 

[PeterW]

It’s to do with getting a decent thermosyphon as a Willis in a tank situation uses the cold of the header tank and the rising water temperature in the feed to create the thermosyphon effect.

 

In this instance it’s only on when the pump is on so doesn’t require the thermosyphon to work. 

 

The other factor is thermostat failure on the immersion so it could boil however most modern thermostats have a secondary non resettable fail stat at 80c. 

[Siiimon]

Great. Thanks for your help all. 

Edited September 24, 2018 by Siiimon

[Ed Davies]

BTW, do these Willis heaters, and the like, use standard immersions as fitted to cylinders?

[Jeremy Harris]

1 hour ago, Ed Davies said:

BTW, do these Willis heaters, and the like, use standard immersions as fitted to cylinders?

 

 

Yes, as far as I know.  I've only seen one for real and that was a Willis made one, and it looked to me as if the element was a standard one with a 2 1/4" BSP thread.

[PeterW]

They are a standard 11” immersion in a threaded case - some are pretty much a 63mm copper pipe section with a brazed in end and a brazed brass immersion flange with a couple of 22mm compression fittings attached.  

[Ed Davies]

Thanks JSH and Peter, had long thought of making something similar myself (for PV diversion into big thermal store) but it'd be a lot better to use an off-the-shelf version now I know they exist (knew about the Willis thermosyphon for solar thermal but for some reason hadn't come across the electric version until recently).

[Siiimon]

Looking at the power connectors for the Willis, how have people switched the unit? I am using a Polypipe system with a master control unit . This looks to be designed for gas or oil fired boilers as heat sources. I can't find any specs on the rating for the boiler contact, so am a bit nervous about connecting a 3kw Willis directly to the rail. How have others approached connecting the Willis to switching?

[PeterW]

1 hour ago, Siiimon said:

Looking at the power connectors for the Willis, how have people switched the unit? I am using a Polypipe system with a master control unit . This looks to be designed for gas or oil fired boilers as heat sources. I can't find any specs on the rating for the boiler contact, so am a bit nervous about connecting a 3kw Willis directly to the rail. How have others approached connecting the Willis to switching?

 

You would melt the controller ..!! It needs a contactor in the circuit to handle the high load side of it - £15-20 on eBay or other sites. If you just want to have a play about I think a standard immersion timer would be fine. 

[Dudda]

Have this in my house. Have it on a standard immersion timer currently as it's not long in and still slowly raising the temperature of the slab. Haven't needed it much yet as the heating season is only starting. Plan to connect it to a thermostat on the wall in the soon somehow. How would I go about connecting this to a wall thermostat? I've a three core and earth and a cat-6 cable going to where I want the wall stat in the main living area. Want something simple and not overly complicated to go with the simple heating element.

[Nickfromwales]

1 hour ago, Dudda said:

Have this in my house. Have it on a standard immersion timer currently as it's not long in and still slowly raising the temperature of the slab. Haven't needed it much yet as the heating season is only starting. Plan to connect it to a thermostat on the wall in the soon somehow. How would I go about connecting this to a wall thermostat? I've a three core and earth and a cat-6 cable going to where I want the wall stat in the main living area. Want something simple and not overly complicated to go with the simple heating element.

For reliability I’d go for a solid state relay to switch the Willis heaters. One relay per Willis. 

[PeterW]

1 hour ago, Nickfromwales said:

For reliability I’d go for a solid state relay to switch the Willis heaters. One relay per Willis. 

 

Hmm ... let’s agree to disagree on that one..! 

[Nickfromwales]

36 minutes ago, PeterW said:

 

Hmm ... let’s agree to disagree on that one..! 

You like the arc’ing and clonking of the contactors ? 

Electormechanical vs solid state ? ?

[PeterW]

@Nickfromwales Solid state failure (esp Chinese cr@p) will fail unseen and has no physical indication of either failure or actuation. They are also single pole and leave the neutral in place so cannot be deemed to completely isolate the circuit. 

 

Given it’s likely that these will change over at significantly low frequency, the need to use an SSR that is designed for high speed and high frequency changeover, usually driven at logic level, then the argument for using them is drastically diminished ....

 

your turn .... ?

[Nickfromwales]

First point is moot as single pole switching in a central heating system is always either upsteam or downstream of suitable means of double pole ( dedicated ) isolation. Your precious contactor will, in normal service, be fed by a single pole MCB and not even that will be seen as routine isolation as a well designed system will not see you going into the fuse board to isolate for servicing. MCB are fault devices NOT isolators. Most MIs will call for such equipment to have “a means of LOCAL double pole isolation with a contact separation of at least 3mm” so the solid state relay would first be fed from either a 13a DP fused spur or a 20a DP switch which would be identified as the equipments means of local isolation. 

 

Are you saying a SS relay wouldn’t do the job? Without the clunk?

 

A brief tutorial and please see the section on Solid State Relay Output Circuit where the primary example scenario is based on switching heating loads. 

 

Game, set and match.....”New balls please” ?

[dpmiller]

I'd say that in the kit I work on (autoclaves, incubators) it doesn't matter whether it's SSR or electromechanical, they still fail. Also need to consider a chunk of space for the heatsink an SSR needs to make it reliable.

At 13A non-inductive I'd go with a cheapo plugin relay and base though rather than a full-on contactor.

[PeterW]

Not exactly game set and match as you quote a single item as being a benefit and as @dpmiller says, to maintain reliability you need a honking big heat sink so your pretty SSR is now taking up the space for a single SSR that a 3 pole contactor would use ...  ?

 

As you blatantly ignored the logic level switching argument, the notion of using AC switching at load line level requires bridge rectification and other components within the SSR to enable the activation of the load which further adds to the MTBF reducing argument due to infrared component complexity ....

 

Now where is the umpire ...??? @JSHarris..!!

[Nickfromwales]

16 minutes ago, PeterW said:

Now where is the umpire ...???

 

43 minutes ago, dpmiller said:

At 13A non-inductive I'd go with a cheapo plugin relay and base though rather than a full-on contactor.

There I think  

[Pete]

5 hours ago, Nickfromwales said:

First point is moot as single pole switching in a central heating system is always either upsteam or downstream of suitable means of double pole ( dedicated ) isolation. Your precious contactor will, in normal service, be fed by a single pole MCB and not even that will be seen as routine isolation as a well designed system will not see you going into the fuse board to isolate for servicing. MCB are fault devices NOT isolators. Most MIs will call for such equipment to have “a means of LOCAL double pole isolation with a contact separation of at least 3mm” so the solid state relay would first be fed from either a 13a DP fused spur or a 20a DP switch which would be identified as the equipments means of local isolation. 

 

Are you saying a SS relay wouldn’t do the job? Without the clunk?

 

A brief tutorial and please see the section on Solid State Relay Output Circuit where the primary example scenario is based on switching heating loads. 

 

Game, set and match.....”New balls please” ?

Is that the simple explanation!!

[Jeremy Harris]

FWIW, I've had a 25 A rated SSR switching our equivalent to an immersion heater on and off reliably for a couple of years or so.  The heat sink needed is modest, and the alloy case it's screwed to does a good enough job.  The big advantage is logic level control in my case, as I switch the SSR directly via a microcontroller pin.

 

For a simple application like switching an immersion from an existing low power programmer/thermostat I'd just use a relay, though, as @dpmiller suggests.  You can make a neat job of this by using a small DIN rail case and fitting a plug-in relay that fits a DIN mounting socket base.  The most compact case with a DIN rail is probably one of the ones intended for use with a DP isolator.  You used to be able to buy these on their own cheaply, but a quick web search has failed to come up with one.  I've used the small Wylex ones for jobs like this, but can't for the life of me find the same ones now, which is a shame.

[Ed Davies]

I'm with @JSHarris on this; the mechanical mounting considerations and source of the control signal are more important than any pros/cons of relays vs SSRs for an application like this where switching is not going to happen more than a few times a day.

 

If it was likely to be switching more quickly (e.g., multiple small immersions being switched on and off to approximately follow the changing output of PV as I have in mind) then SSRs would be preferable.