Home made Heat Bank

198 views
Skip to first unread message

robgraham

unread,
Mar 15, 2007, 9:19:45 AM3/15/07
to
Because I live in a cottage with the corresponding low head for the
DHW, I've been pondering for some time on the idea of converting the
hotwater tank to a Heat Bank. Before anyone jumps in with the pump
suggestion, I am under pressure from SWMBO to supply hot water to a
shower and 2 taps at a considerably better flow than at moment - eg
flow at mains pressure would be liked!

I've been reading the DIY Wiki entry on this subject and realise that
stratification control is an issue and would appreciate any input from
anyone who has undertaken this as a project.

A couple of questions as starters :

1 I've seen heat exchangers for combis advertised on Ebay - are
these the sort of heat exchanger one would use ?
2 I take the tank is left vented to the atmosphere via a header tank
but that the working fluid is a closed loop viz. an unpressurised CH
system.

Any comments would be appreciated

Rob

John Stumbles

unread,
Mar 15, 2007, 9:37:29 AM3/15/07
to
On Thu, 15 Mar 2007 06:19:45 -0700, robgraham wrote:

> Because I live in a cottage with the corresponding low head for the
> DHW, I've been pondering for some time on the idea of converting the
> hotwater tank to a Heat Bank. Before anyone jumps in with the pump
> suggestion, I am under pressure from SWMBO to supply hot water to a
> shower and 2 taps at a considerably better flow than at moment - eg
> flow at mains pressure would be liked!
>
> I've been reading the DIY Wiki entry on this subject and realise that
> stratification control is an issue and would appreciate any input from
> anyone who has undertaken this as a project.

That's me :-)

> A couple of questions as starters :
>
> 1 I've seen heat exchangers for combis advertised on Ebay - are
> these the sort of heat exchanger one would use ?

Probably not because (a) they're a bit on the low heat output side (b)
they're intended for connection via proprietary mechanisms involving O
rings and clamps and are a PITA to make DIY connections to.
You can buy them from DPS - as used in their Pandora heat banks, cost
c.£150 IIRC or from GEA Ecoflex. I got a Copper brazed plate heat
exchanger model:- BP12M-20-G1G1 Packed and delivered. £100 + VAT. It
works pretty well but not perfectly: at full flow from the bath tap the
water comes out cool as there's evidently not enough heat transfer at the
primary flow rate I've got.

> 2 I take the tank is left vented to the atmosphere via a header tank
> but that the working fluid is a closed loop viz. an unpressurised CH
> system.

That's how mine is arranged. It means you've got very good recovery as the
primary water from the boiler goes straight into the tank to be drawn on
by the PHE to deliver more hot water, so in practice the supply never runs
cold for long: if you do run it cold you can get hot water again in a few
minutes providing the boiler is running.

When I first tested it with just the electric immersion heater it seemed
to run cold quite quickly and take a long time to recover, which is what
you'd expect.

I've been meaning to do more methodical tests on it and to properly
calculate the spec for the PHE needed, but apart from the Tuits I lack
suitable measuring kit such as flow meters and multiple temperature
sensors for the practical tests, and don't have all the data I need for
the calculations.

Probably the first available Tuit should go to writing it up on the Wiki
so others can see and hopefully contribute.


Andy Hall

unread,
Mar 15, 2007, 9:42:23 AM3/15/07
to
On 2007-03-15 13:19:45 +0000, "robgraham" <robkg...@btinternet.com> said:

> Because I live in a cottage with the corresponding low head for the
> DHW, I've been pondering for some time on the idea of converting the
> hotwater tank to a Heat Bank. Before anyone jumps in with the pump
> suggestion, I am under pressure from SWMBO to supply hot water to a
> shower and 2 taps at a considerably better flow than at moment - eg
> flow at mains pressure would be liked!
>
> I've been reading the DIY Wiki entry on this subject and realise that
> stratification control is an issue and would appreciate any input from
> anyone who has undertaken this as a project.
>
> A couple of questions as starters :
>
> 1 I've seen heat exchangers for combis advertised on Ebay - are
> these the sort of heat exchanger one would use ?

You could, but GEA Ecobraze make some larger ones with more transfer
capacity. In a combi, there may well be a space limitation but then
in typical
domestic appliance sizes, they don't need to be able to transfer more than
40kW and produce greater than a 35 degree temperature rise at the rated flow
rate.

In a heatbank setup, you can easily go for something more substantial
that will give
much better flow rates because more heat can be transfered. 100kW or even more
is possible and the exchanger is still not a great deal larger than a
couple of bricks.

http://www.gea-ecobraze.se/

They also have a UK office in Sheffield who will supply details of suppliers.


> 2 I take the tank is left vented to the atmosphere via a header tank
> but that the working fluid is a closed loop viz. an unpressurised CH
> system.

That's one way to do it, although not all boilers are happy to work on vented
primaries.

Another approach is to keep the bulk hot water in the cylinder as
unvented (perhaps
with a header tank built on the top - the Heatweb ones are like this)
and heat the cylinder
with a coil from the primary circuit which can then be pressurised.

One caveat here is to make sure that the cylinder is a fast recovery
type with plenty of coil surface
area. The traditional and basic ones will work, but are limited in
terms of heat transfer rate.
Ideally one wants to get all of the output of the boiler beginning to
replensih the cylinder as heat is
used.

A further point is to heavily insulate the cylinder. It will be
running at 82 degrees or so rather than the
typical 60 of DHW.


robgraham

unread,
Mar 15, 2007, 6:02:54 PM3/15/07
to
On 15 Mar, 13:42, Andy Hall <a...@hall.nospam> wrote:

OK guys thanks for the 'starters for ten'.

This is a retirement project so cost has to be minimal and one trade
off is that I have an 'standard' DHW tank and a similar one unused.
Neither are likely to be 'fast recovery' type and I'm not sure that
that in a two person house that that is going to be that critical.
I'll do the sums on filling a bath for instance but if the current
tank at 60 C fills the bath perfectly adequately without compromising
hot water elsewhere in the house then surely a tank at 80C is going to
have more reserve. Having seen the quoted transfer ratio across heat
exchangers, that is not going to be the area of any loss.

Insulation doesn't present a problem. Box the tank in and fill with
something (?).

The reason for asking about header tank, etc is the problem of head
room. Currently the DHW tank has 1m head (top of DHW tank to surface
of CW tank) - if I place the hot tank on the attic floor that headroom
is going to disappear. That would push me down the 'pressurised'
primary route but that does assume that the boiler will cope with that
- and equally the CH system (!). Apart from the obvious (ie the water
is under pressure!) what Is the difference between systems that will
cope with pressure and those that don't ?

Thanks

Rob

Andy Hall

unread,
Mar 15, 2007, 6:46:45 PM3/15/07
to
On 2007-03-15 22:02:54 +0000, "robgraham" <robkg...@btinternet.com> said:
>
> OK guys thanks for the 'starters for ten'.
>
> This is a retirement project so cost has to be minimal and one trade
> off is that I have an 'standard' DHW tank and a similar one unused.
> Neither are likely to be 'fast recovery' type and I'm not sure that
> that in a two person house that that is going to be that critical.

Probably not.

> I'll do the sums on filling a bath for instance but if the current
> tank at 60 C fills the bath perfectly adequately without compromising
> hot water elsewhere in the house then surely a tank at 80C is going to
> have more reserve.

Very roughly, you can start on the basis of 4/3 of the volume that you would
get at 60 degrees. However, it may be worse because of circulation
of the water in
the cylinder tending to stir it.


> Having seen the quoted transfer ratio across heat
> exchangers, that is not going to be the area of any loss.

Yes and no but be careful. It is better to have a heat exchanger
that is oversized and not let it
be the limiting factor. You can control HW temperature by using a
thermostatic blending valve
to handle changes of CW temperature and flow rates.

Just on that point..... the standard health warning about checking for
adequate mains pressure and flow before starting.


>
> Insulation doesn't present a problem. Box the tank in and fill with
> something (?).

Reasonable. Rockwool or glass fibre, or if space is tight, Celotex.
You could locate the heat exchanger inside the box as well.

>
> The reason for asking about header tank, etc is the problem of head
> room. Currently the DHW tank has 1m head (top of DHW tank to surface
> of CW tank) - if I place the hot tank on the attic floor that headroom
> is going to disappear. That would push me down the 'pressurised'
> primary route but that does assume that the boiler will cope with that
> - and equally the CH system (!). Apart from the obvious (ie the water
> is under pressure!) what Is the difference between systems that will
> cope with pressure and those that don't ?

In a lot of ways, not a great deal. Have a look at Ed Sirett's sealed
system FAQ.

Are you sure that the cylinders you have are direct types? Otherwise
you will have in any case the bulk water in the cylinder unvented and a
choice between vented and sealed for the primary depending on what you
want and what the boiler can do.

For the cylinder, if you have a coil and the water is therefore
separate to the primary, the arrangement can be easy. If you look at
the Heatweb units, they have a small header tank on the top a bit like
a Fortic tank.
You could easily create something similar by making a frame for the
cylinder and extending it upwards, putting the small header tank on the
top.

If you have a coil and hence a separate primary, open vented, then
nothing much changes if that's what you have now. The header tank for
this can even be on the floor of the attic if you want.

If you have a direct cylinder and the water is common between cylinder
and heating, then you need to be more careful about pump,
feed/expansion pipe and vent pipe siting in order to make sure that
there is not pumping over, because now this header tank will have to be
above the cylinder and the head will be small.

Antony Jones

unread,
Mar 15, 2007, 7:01:08 PM3/15/07
to

Stand the cylinder on some rigid insulation, say 4" of polystyrene or
Celotex if you really want. Surround the cylinder with 8" of glass wool
- it's dirt cheap and will fill the gap nicely between the cylinder and
the (presumably) square box. Put plenty on top of the cylinder too, and
have a tight-fitting lid to stop convection.

> The reason for asking about header tank, etc is the problem of head
> room. Currently the DHW tank has 1m head (top of DHW tank to surface
> of CW tank) - if I place the hot tank on the attic floor that headroom
> is going to disappear. That would push me down the 'pressurised'
> primary route but that does assume that the boiler will cope with that
> - and equally the CH system (!). Apart from the obvious (ie the water
> is under pressure!) what Is the difference between systems that will
> cope with pressure and those that don't ?

I'm confused by this - do you mean the DHW tank or the F&E tank? My
thermal store (an Albion Mainsflow Direct) sits in the loft on the
"raft" that used to support the 50 gallon cold water tank. The Mainsflow
has it's own float valve at the top, so it's also the F&E tank for
itself and the primary circuits (the boiler and radiators are two
separate circuits with their own pumps).

For your purposes, you could put the heatbank in the loft and use a
pumped supply from the water tank to the heat exchanger. A small (4
gallon) F&E tank could be placed on top of your box to keep the
heatbank/primary circuit topped up.

Or you could do away with the water tank and go for a mains pressure
feed direct to the heat exchanger. The heatbank/primary circuit would
still be under low pressure.

Or you could make the heatbank indirect and have a sealed primary
circuit (the heatbank could then be filled with a hose, like the DPS
Pandora).

I'd suggest you don't try to make the heatbank itself pressurised as
that requires extra safety devices like an unvented cylinder has.

-Antony.

robgraham

unread,
Mar 16, 2007, 7:31:25 AM3/16/07
to
<snip>

>
> I'm confused by this - do you mean the DHW tank or the F&E tank? My
> thermal store (an Albion Mainsflow Direct) sits in the loft on the
> "raft" that used to support the 50 gallon cold water tank. The Mainsflow
> has it's own float valve at the top, so it's also the F&E tank for
> itself and the primary circuits (the boiler and radiators are two
> separate circuits with their own pumps).
>
> For your purposes, you could put the heatbank in the loft and use a
> pumped supply from the water tank to the heat exchanger. A small (4
> gallon) F&E tank could be placed on top of your box to keep the
> heatbank/primary circuit topped up.
>
> Or you could do away with the water tank and go for a mains pressure
> feed direct to the heat exchanger. The heatbank/primary circuit would
> still be under low pressure.
>
> Or you could make the heatbank indirect and have a sealed primary
> circuit (the heatbank could then be filled with a hose, like the DPS
> Pandora).
>
> I'd suggest you don't try to make the heatbank itself pressurised as
> that requires extra safety devices like an unvented cylinder has.
>
> -Antony.

Yes - I got myself a bit confused !!

As I see it there is 3 circuits -

1 The primary circuit to/from the boiler. In my case just to
complicate matters, I have a wood burner included and a Dunsley
Neutraliser is used to blend the two heat sources -interestingly I
saw someone once suggest that the Dunsley was in fact a miniature
heatbank. Anyway that's all vented with a header tank and a pump for
the CH and a second from the oil burner.

2 Then we have the heatbank/pump/heat exchanger circuit. The tank,
being an ordinary one, cannot be at pressure, so I need a second
header tank for that at atmostpheric pressure !

3 Not really a 'circuit' but the path from the mains water through
the heat exchanger to the shower and taps.

One downer is that I suspect I will have to upgrade the feed from the
riser from 15mm to 22mm. It's something like a 15m run to the attic.
Could someone say ya or nay to that for me please. That could a
significant problem to tackle - can that be done in plastic for
instance ?

Thanks for all your answers

Rob

Andy Hall

unread,
Mar 16, 2007, 9:08:55 AM3/16/07
to
On 2007-03-16 11:31:25 +0000, "robgraham" <robkg...@btinternet.com> said:
>
> As I see it there is 3 circuits -

As you describe it, yes.

If the cylinder were a direct one (meaning no coil and the bulk water
in it circulates
to boiler and CH) then it would be one for that and one for the mains side.

>
> 1 The primary circuit to/from the boiler. In my case just to
> complicate matters, I have a wood burner included and a Dunsley
> Neutraliser is used to blend the two heat sources -interestingly I
> saw someone once suggest that the Dunsley was in fact a miniature
> heatbank. Anyway that's all vented with a header tank and a pump for
> the CH and a second from the oil burner.

That's OK. Given that you have all of this, and if it works OK, I
don't see an over-riding
reason to change it to sealed as long as you are using an indirect cylinder.

The other thing that occurs to me is that with the woodburner as well,
you may well
not have the amount of control or output to get the temperature of
cylinder up to 80 degrees.

If the setup can only manage to raise it to 60 degrees as it might now,
you may even be a little
worse off. You would be able to get a greater flow and pressure than
at present, but for a shorter
period of time. That may matter for a shower.


>
> 2 Then we have the heatbank/pump/heat exchanger circuit. The tank,
> being an ordinary one, cannot be at pressure, so I need a second
> header tank for that at atmostpheric pressure !

Exactly, and it needs to be located above the cylinder. The one for
the primary (1) may need to be raised
a bit as well because it if it sitting on the floor of the attic, it
needs to be above the level of the coil in the
cylinder. Realistically, it may make sense to build a platform to take
both header tanks and done with it.


>
> 3 Not really a 'circuit' but the path from the mains water through
> the heat exchanger to the shower and taps.

Exactly.


>
> One downer is that I suspect I will have to upgrade the feed from the
> riser from 15mm to 22mm. It's something like a 15m run to the attic.
> Could someone say ya or nay to that for me please. That could a
> significant problem to tackle - can that be done in plastic for
> instance ?

It should be 22mm otherwise the flow will be quite restricted unless
the pressure is substantial.
Plastic would be OK.


Dave Liquorice

unread,
Mar 16, 2007, 9:08:08 AM3/16/07
to
On 16 Mar 2007 04:31:25 -0700, robgraham wrote:

> 1 The primary circuit to/from the boiler. In my case just to
> complicate matters, I have a wood burner included and a Dunsley
> Neutraliser is used to blend the two heat sources -interestingly I
> saw someone once suggest that the Dunsley was in fact a miniature
> heatbank. Anyway that's all vented with a header tank and a pump for
> the CH and a second from the oil burner.
>
> 2 Then we have the heatbank/pump/heat exchanger circuit. The tank,
> being an ordinary one, cannot be at pressure, so I need a second
> header tank for that at atmostpheric pressure !

I might be (probably am) missing something but what is the function of
this intermediate "circuit". Why can't the boiler/CH water be the same as
that used in the heat bank? Each heat transfer stage must loose
something.

--
Cheers new...@howhill.com
Dave. pam is missing e-mail

robgraham

unread,
Mar 16, 2007, 9:34:35 AM3/16/07
to
On 16 Mar, 13:08, Andy Hall <a...@hall.nospam> wrote:
>(snipped)

>
> The other thing that occurs to me is that with the woodburner as well,
> you may well
> not have the amount of control or output to get the temperature of
> cylinder up to 80 degrees.
>
> If the setup can only manage to raise it to 60 degrees as it might now,
> you may even be a little
> worse off. You would be able to get a greater flow and pressure than
> at present, but for a shorter
> period of time. That may matter for a shower.

I'm sure I can find a way round that with sensors of one sort or
another. There is already a logic box control the system. If I head
off in my box before my wife, heaven help the CH technician if
maintenance is required !

>Realistically, it may make sense to build a platform to take
> both header tanks and done with it.
>

Is there a good reason why one header tank with 2 separate outlets
couldn't be used ?

Thanks Andy

Rob


John Stumbles

unread,
Mar 16, 2007, 10:45:40 AM3/16/07
to
On Thu, 15 Mar 2007 13:37:29 +0000, John Stumbles wrote:

> Probably the first available Tuit should go to writing it up on the Wiki
> so others can see and hopefully contribute.

OK first draught here:
http://wiki.diyfaq.org.uk/index.php?title=DIY_Heat_Bank

Andy Hall

unread,
Mar 16, 2007, 12:02:54 PM3/16/07
to

It can

Rob already has the indirect cylinder though

Of course, something like some Essex flanges could be fitted to
create additional accesses to the main part of the cylinder to make
it a direct one.


Andy Hall

unread,
Mar 16, 2007, 12:05:07 PM3/16/07
to

Then in effect, you will have created a common system. I can't see a
reason why not
but then if you are going to do that, you might as well go for a direct
cylinder approach
by putting a couple of additional connection bosses on the cylinder -
e.g. Essex flanges


John Stumbles

unread,
Mar 16, 2007, 5:05:17 PM3/16/07
to
On Fri, 16 Mar 2007 16:05:07 +0000, Andy Hall wrote:

> Then in effect, you will have created a common system. I can't see a
> reason why not
> but then if you are going to do that, you might as well go for a direct
> cylinder approach
> by putting a couple of additional connection bosses on the cylinder -
> e.g. Essex flanges

Or even just teeing into the top out and bottom feed connections.


Andy Hall

unread,
Mar 16, 2007, 5:34:46 PM3/16/07
to

Mmmm....

Although what's going to happen when water is being drawn and the boiler fires?

You wouldn't want to have one pump sucking and the other blowing at
each connection
because that would effectively circulate water from the boiler directly
through the heat exchanger
and back. In effect, the cylinder wouldn't contribute anything much
and one would have a
crappier than usual combi.

Operationally, I would have thought that one would want to feed hot
water from the boiler to the
top of the cylinder and take cool from the bottom, whereas for the heat
exchanger one would want
to take hot from the top and return cool to the bottom.

If the flows are arranged such that both suck and blow at the same
locations, then hot is going in
from the boiler and cold from the exchanger.


I have another creative idea.


- Get an Abrafile blade and suitably attach it to two cords

- Drop assembly through immersion heater boss and arrange it to loop
around the pipe of
the coil near the entry fitting.

- Operate saw by pulling the ends of the cord until pipe is cut.

- Relocate cord near to lower entry fitting of coil and saw again.

- Manipulate coil out through the immersion heater boss (or treat it
like a channel tunnel drilling machine)

- voila


This exercise may be time consuming. Rather like solving one of those
metal puzzles involving string
and nails.


Stephen Dawson

unread,
Mar 16, 2007, 5:35:57 PM3/16/07
to


"John Stumbles" <john.s...@ntlworld.com> wrote in message
news:pan.2007.03.16....@ntlworld.com...

What about a thermal store ?? We quite like Gledhill Boilermates see
www.gledhill.net fo more info. Mains pressure hot water.

--
Regards

Stephen Dawson
Fox Electrical Services Ltd
www.foxelectrical.co.uk


Andy Hall

unread,
Mar 16, 2007, 5:41:57 PM3/16/07
to
On 2007-03-16 21:35:57 +0000, "Stephen Dawson"
<stephe...@ntlworld.com> said:

>
>
>
> "John Stumbles" <john.s...@ntlworld.com> wrote in message
> news:pan.2007.03.16....@ntlworld.com...
>> On Fri, 16 Mar 2007 16:05:07 +0000, Andy Hall wrote:
>>
>>> Then in effect, you will have created a common system. I can't see a
>>> reason why not
>>> but then if you are going to do that, you might as well go for a direct
>>> cylinder approach
>>> by putting a couple of additional connection bosses on the cylinder -
>>> e.g. Essex flanges
>>
>> Or even just teeing into the top out and bottom feed connections.
>>
>>
>
>
>
> What about a thermal store ?? We quite like Gledhill Boilermates see
> www.gledhill.net fo more info. Mains pressure hot water.

Project using existing materials to hand....


John Stumbles

unread,
Mar 16, 2007, 6:39:01 PM3/16/07
to
On Fri, 16 Mar 2007 21:34:46 +0000, Andy Hall wrote:

> Although what's going to happen when water is being drawn and the boiler
> fires?
>
> You wouldn't want to have one pump sucking and the other blowing at each
> connection
> because that would effectively circulate water from the boiler directly
> through the heat exchanger
> and back. In effect, the cylinder wouldn't contribute anything much and
> one would have a
> crappier than usual combi.

Depends. Assuming the DHW pump has more oomph[1] than the CH one[2] then
it should draw water from the cylinder as well as what the boiler can
supply.


[1] technical term :-)

[2] which might need to be arranged by, say, a gate valve in the
DHW part of the primary circuit if the length of pipework, the boiler
itself and the motorised valve aren't enough.

robgraham

unread,
Mar 16, 2007, 7:31:33 PM3/16/07
to
On 16 Mar, 16:02, Andy Hall <a...@hall.nospam> wrote:

The presumed advantage of the direct system for a heatbank is that is
recharges quicker. I started off this Reply thinking that I would
have to stick with an indirect system, but in fact the Neutraliser set-
up I have may actually make the direct system work perfectly well as
it would act as a buffer between the tank and the CH system. Need to
think about that. I like the concept of the abra file saws ! I take
it that once the coil is in, it can't be got out again ?

On the subject of the Essex Flange, should the take off be through one
of these ? I have a feeling I've seen that it should be to preserve
the stratification.

Rob

John Stumbles

unread,
Mar 17, 2007, 4:08:50 AM3/17/07
to
On Fri, 16 Mar 2007 16:31:33 -0700, robgraham wrote:

> The presumed advantage of the direct system for a heatbank is that is
> recharges quicker. I started off this Reply thinking that I would
> have to stick with an indirect system, but in fact the Neutraliser set-
> up I have may actually make the direct system work perfectly well as
> it would act as a buffer between the tank and the CH system.

AIUI the neutraliser is like a mini heat bank so if you were prepared to
add tappings to your tank you could probably bung both your heat sources
into it directly and do away with the neutraliser. OTOH since you've got
it already them if that bit's working OK, I'd just keep it.

Need to
> think about that. I like the concept of the abra file saws ! I take
> it that once the coil is in, it can't be got out again ?

I think Andy was talking about taking the coil out with the abrafile, and
you certainly wouldn't have the option of getting it back in. Given the
choice of trying to cut through two 28mm pipes inside the cylinder with an
abrafile and bunging on a couple of flanges I'd go for the latter!


> On the subject of the Essex Flange, should the take off be through one
> of these ? I have a feeling I've seen that it should be to preserve
> the stratification.

You're going to bugger up the stratification pdq by pumping water from the
cylinder through a PHE, but I think the quick recovery of having the
cylinder directly heated makes up for this. I used the tappings I did on
my direct cylinder because they were there but I think you could equally
well just use the top and bottom tappings. If I had a few Tuits (and SWMBO
hadn't filled the attic with so much junk I can't get near the system) I'd
change over the connections on my own system and try it that way and let
you know.

Andy Hall

unread,
Mar 17, 2007, 5:50:41 AM3/17/07
to
On 2007-03-16 23:31:33 +0000, "robgraham" <robkg...@btinternet.com> said:

>>
>
> The presumed advantage of the direct system for a heatbank is that is
> recharges quicker.

The point is that it in effect it is accepting all of the heat that the
boiler can
manage. When there is a coil, it is acting as a heat exchanger itself and
the heat transfer depends on the flow rate, the surface area and the
temperature
difference between the water inside the cylinder and what's going through the
coil. The fast recovery cylinders have coils with larger surface
area - e.g. by using a bundle
of smaller pipes.

> I started off this Reply thinking that I would
> have to stick with an indirect system, but in fact the Neutraliser set-
> up I have may actually make the direct system work perfectly well as
> it would act as a buffer between the tank and the CH system. Need to
> think about that.

Effectively it's a buffer between the stove and the boiler as heat
sources, AIUI.

> I like the concept of the abra file saws ! I take
> it that once the coil is in, it can't be got out again ?

I would think that during manufacture, th coil goes in before the bottom of the
cylinder is attached.

Whether one could feed it out of the immersion heater boss I am not
sure, but it
doesn't really matter if it can't be.

The idea was to use the bosses of the coil to connect the boiler and
make a direct cylinder.


>
> On the subject of the Essex Flange, should the take off be through one
> of these ? I have a feeling I've seen that it should be to preserve
> the stratification.


Again this was a way to add connections to the cylinder for the boiler.


Owain

unread,
Mar 17, 2007, 8:20:58 AM3/17/07
to
John Stumbles wrote:
> Depends. Assuming the DHW pump has more oomph[1]
> [1] technical term :-)

oomph is to pumps as whoosh is to sewer falls.

Owain

John Rumm

unread,
Mar 17, 2007, 10:59:20 AM3/17/07
to
Andy Hall wrote:

>> This is a retirement project so cost has to be minimal and one trade
>> off is that I have an 'standard' DHW tank and a similar one unused.
>> Neither are likely to be 'fast recovery' type and I'm not sure that
>> that in a two person house that that is going to be that critical.
>
> Probably not.

If you are running a vented primary, you could also convert the cylinder
to a direct one by the addition of a couple of Essex flanges, and just
ignore the indirect coil.


--
Cheers,

John.

/=================================================================\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\=================================================================/

Andy Hall

unread,
Mar 17, 2007, 11:16:19 AM3/17/07
to

You mean if there isn't enough then a big smell ensues?


John Rumm

unread,
Mar 17, 2007, 11:14:36 AM3/17/07
to
John Stumbles wrote:

>> Probably the first available Tuit should go to writing it up on the Wiki
>> so others can see and hopefully contribute.
>
> OK first draught here:
> http://wiki.diyfaq.org.uk/index.php?title=DIY_Heat_Bank

Cute, the TMV is a nice touch often forgotten that ought to make the
system more user friendly.

On a slightly different topic, what would you say is the best way to
mitigate the conflicting requirements[1] for boiler flow temperature
when using a modern boiler in a system like this? Just balance the CH
for a larger temperature drop?

[1] i.e. for optimal boiler efficiency the lowest adequate flow return
temperature is desirable to promote condensing, but the heat bank will
benefit from a higher storage temp.

Andy Hall

unread,
Mar 17, 2007, 12:24:28 PM3/17/07
to
On 2007-03-17 15:14:36 +0000, John Rumm <see.my.s...@nowhere.null> said:

> John Stumbles wrote:
>
>>> Probably the first available Tuit should go to writing it up on the Wiki
>>> so others can see and hopefully contribute.
>>
>> OK first draught here:
>> http://wiki.diyfaq.org.uk/index.php?title=DIY_Heat_Bank
>
> Cute, the TMV is a nice touch often forgotten that ought to make the
> system more user friendly.
>
> On a slightly different topic, what would you say is the best way to
> mitigate the conflicting requirements[1] for boiler flow temperature
> when using a modern boiler in a system like this? Just balance the CH
> for a larger temperature drop?
>
> [1] i.e. for optimal boiler efficiency the lowest adequate flow return
> temperature is desirable to promote condensing, but the heat bank will
> benefit from a higher storage temp.

I've had this debate on numerous occasions with Drivel.

The discussion is somewhat circular, though.

Assuming that the boiler comes on when either the cylinder is cold or
virtually so...

In that scenario, the cold water from the cylinder will flow to the
boiler and it will
heavily condense, with water flowing out to the cylinder determined by
the flow rate, temperature rise, efficiency of the boiler heat
exchanger (high) and the
burner output (max).

As the water warms, the temperatures will rise and the boiler become
less efficient.

This is also true if the cylinder is indirect because now the heat
transfer rate from the
coil to the water in the cylinder is influenced by the transfer through
the coil. If anything the
effect will happen earlier

If the target temperature for the cylinder is only 60 degrees (i.e. use
of HW from there), then
the period of time when the boiler is running at higher temperatures is
less than it is
if the target temperature is 80 degrees.

There is a red herring here which says that if the store temperature is
maintained at 74 degrees,
and the boiler heat exchanger can operate with a 20 degree temperature
change, then the boiler
is always condensing. The argument is bogus because all that happens
when condensing is going
on is an increase in the *rate* of efficiency with falling temperature.
There isn't a Holy Grail.

Turning down the store thermostat will obviously improve boiler
efficiency when running. Where
there is an indirect coil it will reduce cycling at the top end -
because as the cylinder temperature
approaches the boiler max temperature, the boiler will cycle if it's a
simple one or modulate otherwise
increasing the time for full recovery. Of course the latter may not
matter because the point of the store
is that it is storing energy at a temperature above that required for use.


Most condensing boilers can't tell the difference between being
presented with a heating load that is a set of cold radiators as
opposed to a cold cylinder - i.e. they are not given any signal that
tells them what they are supposed
to be doing. For CH, one wants the boiler to modulate down to match
the house heat loss at the
outside temperature. For HW (thermal store or DHW from cylinder) one
wants the boiler to go flat bollock
out until the target temperature is reached.

John Rumm

unread,
Mar 17, 2007, 1:33:26 PM3/17/07
to
Andy Hall wrote:

>> On a slightly different topic, what would you say is the best way to
>> mitigate the conflicting requirements[1] for boiler flow temperature
>> when using a modern boiler in a system like this? Just balance the CH
>> for a larger temperature drop?
>>
>> [1] i.e. for optimal boiler efficiency the lowest adequate flow return
>> temperature is desirable to promote condensing, but the heat bank will
>> benefit from a higher storage temp.
>
> I've had this debate on numerous occasions with Drivel.

Sorry, I think we are talking at cross purposes slightly... (and dribble
was usually advocating sticking a store between boiler and rads as well
thereby knackering its ability to load match)

I was not particularly worried about efficiency of heating the store -
as you say, for all but the last few degrees you will be close to
optimal anyway. I was thinking more about the CH role. Since many/most
boilers don't have a way of externally influencing the flow temperature,
you end up running your rads on a higher flow temperature that you
otherwise might in order to be able to achieve the desired store
temperature. (Hence associated touch risk from the rads, and also the
potential that you lose some efficiency from the boiler while running
the heating steady state, modulated down. You also would potentially get
slightly shorter cycling due to the higher rad output at the 80 degree
flow temp).

> Assuming that the boiler comes on when either the cylinder is cold or
> virtually so...

<snip, lots of "accurate, but not really relevant to my question" stuff
on store reheat ;-) >

> to be doing. For CH, one wants the boiler to modulate down to match
> the house heat loss at the
> outside temperature.

Hence my question really. Obviously it partly depends on how the boiler
does its load matching - but one presumes that return temp, and selected
flow temp will be the major inputs into the control loop. It would be
nice to have a way of automatically selecting a lower flow temp when
running the rads. I can see that balancing the system to drop more per
pass will aid the efficiency of the boiler (while compromising the
maximum available output a little), but does not reduce the maximum
touch temperature that much. Alternatively you could balance for less CH
circuit drop, and rely on the load matching to reduce boiler power more...

Andy Hall

unread,
Mar 17, 2007, 2:39:40 PM3/17/07
to

Now I see what you're saying, John.

Assuming a modulating boiler with no input signals other than one
demand on/off, flow temp control and temperature sensors on return and
flow and controlling only burner rate (not pump speed), the options are
more limited than where there is more instrumentation and control.

However, are we sure that the thermostat setting is what the boiler
will always strive to do?

For large output requirements and cold water coming in, (i.e. startup)
the burner can be going flat out but the
flow will still only be 20-25 degrees above return temp.

However, I am wondering what it will do where there are TRVs involved.
As the house warms up, the TRVs begin to close and flow reduces. Some
will turn off. Less heat is being dissipated by the radiators and if
the return temperature is being maintained as low as possible, the
boiler will presumably throttle back to prevent the flow
temp rising too high. The question is whether the boiler will attempt
to maintain the flow temperature at (say) 80 degrees or will allow it
to drop since there is enough heat being produced.

Apart from the possible safety issue of radiators running at high
temperatures (although that can be looked on as a reason to make the
radiator covers that you've always wanted); if the boiler could work in
this way, it ought to be able to figure out the difference in
characteristics of the HW load vs the CH load. In ideal
circumstances, where the water in the store is not being slooshed
around (technical term) by the draw off through the heat exchanger, the
cold water from the bottom should be presented to the boiler until
quite late in the reheating cycle.

On my boiler, there are a lot of inputs and controlled outputs. Apart
from those mentioned above, there are:

- Sensor on cylinder providing boiler with means to calculate actual
temperature rather than whether or not
heat is needed

- Room thermostat which provides boiler with actual room temperature data

- Outside temperature sensor

- Continuous control of pump speed

- Control of DHW and CH zone valves separately.

- Readout for all temperature sensors, pump %age and fan speed
(proportional to burn rate)

I can also hook up a PC and see exactly what is going on

With all of that configured, the DHW temperature is set on the boiler
and there are also max settings for flow temperature in DHW mode and CH
mode (I have them set to 85 and 70 degrees respectively.

In CH mode, it's clear that the boiler attempts to minimise the return
temperature. It certainly doesn't attempt to maintain a high flow
temperature explicitly AFAICS. During mild weather, it can be
running at 40 degrees flow and not much more than 20 return, with about
4kW output and the pump at about 30% of max.

When DHW reheat is required, the zone valves are switched by the
boiler, to connect to the cylinder and both burner and pump wind up to
full tilt. I can watch the cylinder temperature and boiler return
temperature on a graph on the PC. As the cylinder temperature gets
to about 57 degrees, the boiler starts to modulate back and then power
drops gradually. There is no overshoot of cylinder temperature.

As an experiment, I tried replacing the room thermostat (which is a
special one) with a standard one and the cylinder thermostat likewise.
I configured the outside sensor off as well. In effect this is giving
it almost the same control sensors that a typical boiler would have
apart from knowing the difference between CH and DHW load and being
able to control the pump output.

There is some observed difference in behaviour.

In CH mode, there is still modulation down such that the boiler still
does not attempt to maintain the set flow temperature. However, the
control is not as accurate and there is more of a tendency to cycling
at lower outputs.
Also, the early morning startup and behaviour in the winter when
outside temperatures drop rapidly (e.g. sky clears and it gets dark) is
not as good. Nonetheless, it is fairly clear from comparing the
data on the PC that the boiler still makes a reasonable attempt to
maintain low operating temperatures even if based only on the flow and
return sensors.

In DHW mode, the behaviour is more obviously different. The boiler
does still begin to modulate back, but nothing like as much as with
cylinder temperature sensing and much closer to the end of the reheat
cycle. The cylinder temperature will overshoot by about 5 degrees
because of the detection by the cylinder thermostat (bimetallic in the
test).

So the conclusion is, I think, that it is *possible* that a more basic
boiler *could* do some reasonably intelligent things based on flow and
return temperature sensing. Whether it does, is another thing.


John Rumm

unread,
Mar 17, 2007, 5:36:47 PM3/17/07
to
Andy Hall wrote:

<interesting analysis of boiler response snipped>

> So the conclusion is, I think, that it is *possible* that a more basic
> boiler *could* do some reasonably intelligent things based on flow and
> return temperature sensing. Whether it does, is another thing.

The reality is that one would need to carry our reasonably detailed
analysis on a case by case basis, and then design suitable external
"influences" to modify the behaviour of the control system. So probably
not much chance of that happening. I must admit I can't think of many
generic solutions to this one, other than specifying a boiler like yours
that can differentiate between CH and HW roles.

Andy Hall

unread,
Mar 17, 2007, 6:05:10 PM3/17/07
to
On 2007-03-17 21:36:47 +0000, John Rumm <see.my.s...@nowhere.null> said:

> Andy Hall wrote:
>
> <interesting analysis of boiler response snipped>
>
>> So the conclusion is, I think, that it is *possible* that a more basic
>> boiler *could* do some reasonably intelligent things based on flow and
>> return temperature sensing. Whether it does, is another thing.
>
> The reality is that one would need to carry our reasonably detailed
> analysis on a case by case basis, and then design suitable external
> "influences" to modify the behaviour of the control system. So probably
> not much chance of that happening. I must admit I can't think of many
> generic solutions to this one, other than specifying a boiler like
> yours that can differentiate between CH and HW roles.

One simple solution that would do at least something, would be to have
some sort of arrangement to control the boiler thermostat to be
different for each role. Something with resistors, a pot and a relay
might do it.

At least one could achieve the goal of buckets of heat for the cylinder
and less for the CH.


John Rumm

unread,
Mar 17, 2007, 7:28:51 PM3/17/07
to
Andy Hall wrote:

> One simple solution that would do at least something, would be to have
> some sort of arrangement to control the boiler thermostat to be
> different for each role. Something with resistors, a pot and a relay
> might do it.

Yup, could work, chances are the temperature knob is just a pot...

perhaps there is a market opportunity here to make a module that "breaks
in" to the wiring loom to insert an external control element into the HW
temp pot. Would obviously be boiler specific, but be something with male
and female connectors that simply is inserted between PCB and the front
panel leads.

> At least one could achieve the goal of buckets of heat for the cylinder
> and less for the CH.

Yup.

If I had a thermal store I might even have given it a try!

(although the sexier solution would be using an ePot on on an I^2C
control bus, hooked up to a microcontroller. Then you could really go to
town)

John Stumbles

unread,
Mar 18, 2007, 9:14:08 PM3/18/07
to
On Sat, 17 Mar 2007 23:28:51 +0000, John Rumm wrote:


> Yup, could work, chances are the temperature knob is just a pot...

But prolly mounted directly on the PCB which is mounted behind the front
panel. (At least it is in some boilers.)

John Stumbles

unread,
Mar 18, 2007, 9:17:54 PM3/18/07
to
On Sat, 17 Mar 2007 21:36:47 +0000, John Rumm wrote:


> The reality is that one would need to carry our reasonably detailed
> analysis on a case by case basis, and then design suitable external
> "influences" to modify the behaviour of the control system. So probably
> not much chance of that happening. I must admit I can't think of many
> generic solutions to this one, other than specifying a boiler like yours
> that can differentiate between CH and HW roles.

IIRC the (now discontinued) Worcester Greenstar HE series had a built-in
diverter valve which allowed it to respond differently when providing DHW
than when doing CH. Possibly something like the current Greenstart CDi
conventional might do the same trick.

John Rumm

unread,
Mar 18, 2007, 9:17:55 PM3/18/07
to

The front panel is a separate module to the PCB on my Isar..

Doctor Drivel

unread,
May 26, 2007, 8:59:21 AM5/26/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:45fc...@nt1.hall.gl...

The above is tripe as dedicated heat banks operate at 74C comfortably and
return low temperatures for most of the re-heat time.

> efficiency when running. Where
> there is an indirect coil it will reduce cycling at the top end - because
> as the cylinder temperature
> approaches the boiler max temperature, the boiler will cycle if it's a
> simple one or modulate otherwise
> increasing the time for full recovery. Of course the latter may not
> matter because the point of the store
> is that it is storing energy at a temperature above that required for use.
>
>
> Most condensing boilers can't tell the difference between being presented
> with a heating load that is a set of cold radiators as opposed to a cold
> cylinder - i.e. they are not given any signal that tells them what they
> are supposed
> to be doing. For CH, one wants the boiler to modulate down to match the
> house heat loss at the
> outside temperature. For HW (thermal store or DHW from cylinder) one
> wants the boiler to go flat bollock
> out until the target temperature is reached.

That can be designed in, external to the boiler controls.

Doctor Drivel

unread,
May 27, 2007, 6:12:36 AM5/27/07
to

"John Rumm" <see.my.s...@nowhere.null> wrote in message
news:45fc26e2$0$8730$ed26...@ptn-nntp-reader02.plus.net...

> Andy Hall wrote:
>
>>> On a slightly different topic, what would you say is the best way to
>>> mitigate the conflicting requirements[1] for boiler flow temperature
>>> when using a modern boiler in a system like this? Just balance the CH
>>> for a larger temperature drop?
>>>
>>> [1] i.e. for optimal boiler efficiency the lowest adequate flow return
>>> temperature is desirable to promote condensing, but the heat bank will
>>> benefit from a higher storage temp.
>>
>> I've had this debate on numerous occasions with Drivel.
>
> Sorry, I think we are talking at cross purposes slightly... (and dribble
> was usually advocating sticking a store between boiler and rads as well
> thereby knackering its ability to load match)

You haven't a clue about thermal storage that is clear.

> I was not particularly worried about efficiency of heating the store - as
> you say, for all but the last few degrees you will be close to optimal
> anyway. I was thinking more about the CH role. Since many/most boilers
> don't have a way of externally influencing the flow temperature, you end
> up running your rads on a higher flow temperature that you otherwise might
> in order to be able to achieve the desired store temperature.

You use a weather compensator that dictates the CH water section of the
store, keeping the water at the optimum temperature and boiler cycling
eliminated as the water mass ensures this.

> Hence my question really. Obviously it partly depends on how the boiler
> does its load matching - but one presumes that return temp, and selected
> flow temp will be the major inputs into the control loop. It would be nice
> to have a way of automatically selecting a lower flow temp when running
> the rads.

An outside weather compensator does that for you. They pretty cheap these
days.

Doctor Drivel

unread,
May 27, 2007, 6:17:02 AM5/27/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:45fc...@nt1.hall.gl...

The Potterton Promax can have modulation turned off and the bolier just runs
to the flow setpoint, as ina simple boiler. Great for integrated CH/DHW
thermal storage, where the store is centre of the system and controls all.

Doctor Drivel

unread,
May 27, 2007, 6:25:57 AM5/27/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:45fc...@nt1.hall.gl...

Have a "diverter" valve and DHW priority system. The boiler on full temp. A
pipe stat on the CH flow after the valve in the stat circuit, set to a lower
temp for the rads. When on CH, the rads temp are lower. When on DHW full
temp.

Better still install a weather compensator and a DHW priority system.


Andy Hall

unread,
May 27, 2007, 8:48:12 AM5/27/07
to
On 2007-05-27 11:12:36 +0100, "Doctor Drivel" <Min...@nospam.com> said:

>
> "John Rumm" <see.my.s...@nowhere.null> wrote in message
> news:45fc26e2$0$8730$ed26...@ptn-nntp-reader02.plus.net...
>> Andy Hall wrote:
>>
>>>> On a slightly different topic, what would you say is the best way to
>>>> mitigate the conflicting requirements[1] for boiler flow temperature
>>>> when using a modern boiler in a system like this? Just balance the CH
>>>> for a larger temperature drop?
>>>>
>>>> [1] i.e. for optimal boiler efficiency the lowest adequate flow return
>>>> temperature is desirable to promote condensing, but the heat bank will
>>>> benefit from a higher storage temp.
>>>
>>> I've had this debate on numerous occasions with Drivel.
>>
>> Sorry, I think we are talking at cross purposes slightly... (and
>> dribble was usually advocating sticking a store between boiler and rads
>> as well thereby knackering its ability to load match)
>
> You haven't a clue about thermal storage that is clear.

You don't have a clue about the behaviour of control systems.


>
>> I was not particularly worried about efficiency of heating the store -
>> as you say, for all but the last few degrees you will be close to
>> optimal anyway. I was thinking more about the CH role. Since many/most
>> boilers don't have a way of externally influencing the flow
>> temperature, you end up running your rads on a higher flow temperature
>> that you otherwise might in order to be able to achieve the desired
>> store temperature.
>
> You use a weather compensator that dictates the CH water section of the
> store, keeping the water at the optimum temperature and boiler cycling
> eliminated as the water mass ensures this.
>
>> Hence my question really. Obviously it partly depends on how the boiler
>> does its load matching - but one presumes that return temp, and
>> selected flow temp will be the major inputs into the control loop. It
>> would be nice to have a way of automatically selecting a lower flow
>> temp when running the rads.
>
> An outside weather compensator does that for you. They pretty cheap
> these days.

That is only useful if it is integrated with the boiler. The separate
ones operate by cycling the boiler which defeats the object.


Andy Hall

unread,
May 27, 2007, 8:49:18 AM5/27/07
to

Pointless unless inetgrated with the control electronics of the boiler
and able to modulate its burn rate.


Andy Hall

unread,
May 27, 2007, 8:49:46 AM5/27/07
to

Another pointless exercise.


Doctor Drivel

unread,
May 27, 2007, 11:05:35 AM5/27/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:4659...@nt1.hall.gl...

Total nonsense. Better if integrated of course. When the compensator is in
the room stat circuit it works well.


Doctor Drivel

unread,
May 27, 2007, 11:06:54 AM5/27/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:4659...@nt1.hall.gl...

Total tripe.

Doctor Drivel

unread,
May 27, 2007, 11:06:26 AM5/27/07
to

"Andy Hall" <an...@hall.nospam> wrote in message
news:4659...@nt1.hall.gl...

It is clear, as an amateur, you haven't a clue about thermal storage or
boilers.

Reply all
Reply to author
Forward
0 new messages