# Expansion vessel and initial charge pressure.

• posted on September 15, 2003, 1:12 pm
Hi,
I am planning on replacing my existing boiler with a new Worcester Bosch 35CDi (Sealed system).
I have read the FAQ at http://www.makewrite.demon.co.uk/SealedCH.html which is fantastic but I still have a couple of questions if any of you clever folks can help.
Sorry there are so many questions!
A brief summary.
I live in a TALL thin house (5 floors) and the height between the top of the top radiator and the bottom of the boiler is of the order of 15m. I know that I will need an initial system pressure (at the boiler) of at least 1.5bar for this height but how much 'extra' should I allow?
With the above in mind, what is a sensible initial charge pressure for the expansion vessel(s)?
And finally... The system will have a volume of about 130 Litres. I know that the built in expansion vessel will not be big enough but what size vessel (total) do you think I need? (Also, is it worth getting one big external vessel and not using the existing one in the boiler or is it just as good to get a new vessel to add the extra capacity.)
Is there a web site out there which calculates this? or has a set of tables. My guess is that, once all the formulas for e.g. thermal expansion have been entered and tollerances added, it is the sort of calculation that a computer would do very well.
Stephen Rice

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 1:47 pm
With the boiler at the bottom of the installation, you'll need to basically use the highest pressure recommended by the boiler manufacturer. My calculations (which may well be wrong) show a pressure difference of 1.5bar between the top and bottom of the system. This means that the top won't be pressurised at all if you use 1.5bar pressure at the bottom. Unfortunately, some boiler manufacturers give a 1.5bar limit to refilling, so this could cause real problems. They could be mitigated by using a larger expansion vessel (so that the pressure rises less when heated). However, if the boiler manufacturer says 1.5bar is a maximum you shouldn't plan to exceed it.
I can think of two possible solutions if this is the case:
1. Don't install the boiler on the bottom floor. This, of course, might not be possible.
2. Run two primary circuits. One circuit is run from floor 1-3. On floor 3, you have the bottom of another circuit that runs to floor 4 and 5 with its own filling loop, expansion vessel, pressure relief etc. You use a plate heat exchanger between the two circuits and make sure the top circuit's pump only starts when there is a call for heat from upstairs. The top two floors get their own programmable thermostat to run the pump (and zone valve, to prevent gravity circulation).
It might be better to have separate flow and return lines to the primary side of the plate exchanger, so that the top floors can operate when the bottom 3 floors are off. Of course, you could also zone the system further, if required.
I'll try some ASCII art. Only flow lines shown, return lines run adjacent (but don't go though zone valves).
## = 2 port zone valve PT = programmable room thermostat -- = heating flow/return pipes .. = electrical control signal RR = radiator PHE = plate heat exchanger FL = filling loop + other gubbins

FLOOR 5 RR ........ | . . | FLOOR 4 PT +->>----RR . | . +--FL . | FLOOR 3 ##---PHE RR | | | | FLOOR 2 | PT RR | . | | . | FLOOR 1 BOILER->>--+-----##------RR | FL
The boiler interlock (call for heat) is provided by wiring the zone valve microswitches in parallel. This wiring is omitted from the diagram for clarity. The programmable room thermostats drive the zone valve motors directly. This is no need for a separate "whole system" programmer on the heating side. The upstairs thermostat also controls the pump for the upper circuit. If the filling loops are on the 1st and 3rd floors, then pressurise to 1.5bar.
If the hot water cylinder is situated on floor 4 or 5, it could have its own circuit and heat exchanger, or could be piggy backed onto the upstairs circuit with a suitable 3 way or 2 additional zone valves.
The downstairs zone valve can be bypassed with an automatic bypass valve. The radiator in the room with the downstairs programmer should be fitted with lockshield valves both sides to prevent the radiator being turned off.
Further zoning is recommended and can be done easily, if required, particularly on the lower floors. If sub zoning is required for floors 4 and 5, then the room thermostats should control 2 port valves and their microswitches should be used to control the pump and the 2 port valve controlling the plate exchanger.
I hope this all makes sense...
Christian.

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 2:30 pm
Hi Christian,
Thanks for the rapid (and detailed) reply. I loved the ASCII art as well!

basically
1.5bar
Unfortunately,
boiler
I calculate it at 1.5bar too (SI units are great sometimes). I can't find any specific maximum initial system pressure but one of the charts refers to values for a system at 2bar so I assume that they think that this is okay?. Is 0.5 bar enough 'virtual head' for the upper radiators? I hadn't even considered a heat exchanger (nice solution) but it would be great not to have to have the extra complexity if possible.

not
3,
pump
floors
further,
pressurise
own
off.
and
Yes. It does. Very clear indeed. Thank you again.

Have you any advice about expansion vessel size? I have seen another message talk about 10% as a good rule of thumb. I guess this means 10% free space after the space taken up by the water at the filling pressure?
Stephen

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 4:44 pm

In my experience, they are disturbingly small, given how useful they can be.
Christian.

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 4:46 pm
On Mon, 15 Sep 2003 17:28:13 +0100, "news.mistral.net" <sadf> wrote:

It's made by GEA Ecobraze (www.gea-ecobraze.se) and I have a model 18 with 28 plates. They've actually discontinued that specific one now and there are 24 or 30 plate versions.
The size, not including the pipe connections, is 282x127mm and then for a 30 plate version, 71mm deep - in other words, slightly larger than a house brick.
Pipe connections are 22mm and can be made with push on or compression fittings.
I insulated mine by making a box from ply and lining it with Celotex insulation (because I had some offcuts).
The UK subsidiary is called GEA Ecoflex
GEA Ecoflex United Kingdom Ltd. United Kingdom
Tel: +44 1909 551 107 Fax: +44 1909 550 885 E-mail: snipped-for-privacy@gea-ecoflex.com
.andy
To email, substitute .nospam with .gl

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 4:49 pm
That is fantastic. Thank you so much.
I will go and do some browsing...
Stephen

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 15, 2003, 7:22 pm
"news.mistral.net" wrote:

These are tallish floors and many of them? Depending on how you use the house I would be tempted to install two smaller boilers? This will allow for better control of the heating and/or better HW production if your are using combis. Others have said about a secondary heating circuit. If you can get the boiler in the middle floors you can also work things OK.
I will try to look out a boiler instruction manual that quotes the additional expansion volume required.
--
Ed Sirett - Property maintainer and registered gas fitter.
The FAQ for uk.diy is at www.diyfaq.org.uk

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 16, 2003, 8:20 am
Hi Ed,

It is a Georgian house with mostly 3.40m ceilings (apart from the basement :3m and top floor:2.5m(and of course the top of the top radiator is only 1m or so above the floor

The trouble with putting a single (or second) boiler in the middle is that: The house is listed, and getting consent for the flue is a bit tricky. Particularly at the front of the house. Although there are losts of floors, there is only really one room on each floor, at the back of the house, and I am reluctant to have a boiler in it because of the noise and space removal (and from an esthetic view point, because it would stop the room looking as nice!)
(Also, I have already had the 28mm gas pipe run installed (which wasn't all that cheap) and I would love not to have to have another if I can avoid it.). I MUST learn to plan first and do second! (rather than do, do, do, half plan, do, plan properly, do properly!)
I guess, having much the same thoughts as you, a couple of other people (above) have recommended looking at a plate heat exchanger to 'decouple' the floors (sounds too good to be true. Small, light, cheap, easy to plumb. Where is the catch :) ) An advantage of the small number of rooms however is that, not much heat is needed so I guess a heat excahnger should cope.

All that said, if it is not going to be a problem to have a single circuit charged to e.g. 2bar then the benefit of having a simpler system will probably win me over. If you had a chance to note down and recomended expansion sizes, it would be really kind.
Stephen

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 16, 2003, 4:52 pm
On Tue, 16 Sep 2003 09:20:45 +0100, "news.mistral.net" <sadf> wrote:

I was sceptical as well, but the exchanger I am using would be able to transfer 200kW under flow and temperature drop conditions used in typical systems.
I've run a peak of 15kW through it with no problems at all, although normally it's less - it does what it says.
Have you worked out the heating loads for the top of the house? Presumably, these are not going to be hugely more than this?
.andy
To email, substitute .nospam with .gl

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 17, 2003, 8:34 am
Hi Andy ,

As you say, according to the performance figueres, there will be no problem at all.
I also notice from the 'exploded diagram' of the boiler that the DHW (I think) looks like it is heated by a plate heat exchanger (Not a surprise I guess, to keep the mains water zoned from the central heating?) and that must be a much more punishing test.
Thanks again for your help. (Do you read all of these messages? You seam to help everyone. What a good man!)
Stephen

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 16, 2003, 5:42 pm
"news.mistral.net" wrote:

The only down side I can think of that you have to have another fill point, expansion vessel and releif valve and pipework. The balancing of the heat Xchanger v. the primary rads might be fairly crucial.
I've dragged out the instructions for a Vaillant T/max 824e that was sitting in my store room....
...The total system expansion volume should be 10.9% of total system volume if the initial presure is 1.0bar and 15.6% if the initial pressure is 1.5bar.
So for 1.5 bar and 130 litres you need just over 20 litres, less if you break it to two systems.
HTH

--
Ed Sirett - Property maintainer and registered gas fitter.
The FAQ for uk.diy is at www.diyfaq.org.uk

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 17, 2003, 8:27 am
Hi Ed,

What a star! Thanks very much.
I don't suppose you have a 'rule of thumb' figure for the thermal expansion of water between say 10 and 55C?
I have done a 'back of the envelope' calculation and reckon that the working pressure ought to be of the order of 2.1bar with a 20L expansion vessel but I had to guess the thermal expansion of water to be about 2.5% between the fill and working temps. I have found one source of thermal expansion coefficients for water but my physics is now rusty enough that I am not 100% sure (an understatement) that my calculations are right! (they suggest an expansion of about 1%)
Thank you very much again.
Stephen
P.S. My dad would tell me 'Why don't you go and look up the answer yourself' so please, unless you happen to have a table to hand with the info on it, feel free to tell me the same. (It is good for my character!)
Thanks

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 17, 2003, 8:55 am

that
Ah. I just noticed that that was a linear expansion not cubic. So I make it more like 5%. (That seams quite high?)
Stephen

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
• posted on September 18, 2003, 7:31 am
What can I say...
Thank you very much again.
Stephen