Heating system for an ICF house with UFH

[PeterW]

16 minutes ago, Nelliekins said:

And black/green stripy, since we are doing RWH as well!

 

That stuff is extortionate ..!! Just use 50mm green with a wrap of standard black... 

[PeterW]

52 minutes ago, Nelliekins said:

That said I do prefer Speedfit for elbows etc - no idea why but the twist lock seems to give me some comfort

 

Thats the converse with me as I have to remember if I’ve done the lock up ..! I like the fact that with Hep2O it’s push, twist, done...!

 

Same with compression - I only buy decent ones with copper olives and use PTFE paste on them and never had one go on me. 

[Bitpipe]

On 29/12/2018 at 12:35, Onoff said:

 

@Bitpipe's IBC thread:

 

 

 

Ta - was a fun project - not dropped a pump in yet as waiting for sparky to finish putting power down there.

[Nelliekins]

1 hour ago, Bitpipe said:

 

Ta - was a fun project - not dropped a pump in yet as waiting for sparky to finish putting power down there.

 

Looking to copy the IBC part, although we have a window as our secondary means of escape from the basement, as opposed to your door (although a stihl saw and 20 minutes of effort would soon fix that!) 

 

We are having a deck over, and have an egress ladder to get out of the lightwell. There will be a hatch with gas strut assistance as well. 

 

I think your lightwell is bigger than ours, too - we have 4.1m wide by 2.25m front to back, and 2.8m high (7 courses of logix blocks). 

[Bitpipe]

9 hours ago, Nelliekins said:

 

Looking to copy the IBC part, although we have a window as our secondary means of escape from the basement, as opposed to your door (although a stihl saw and 20 minutes of effort would soon fix that!) 

 

We are having a deck over, and have an egress ladder to get out of the lightwell. There will be a hatch with gas strut assistance as well. 

 

I think your lightwell is bigger than ours, too - we have 4.1m wide by 2.25m front to back, and 2.8m high (7 courses of logix blocks). 

 

IBCs are roughly 1m2 so you could squeeze a few in - we put a mezzanine deck over the top of the IBCs which makes a nice feature.

 

TBH we originally envisaged our lightwell / egress area as a tiered construction but the SE pointed out that it would be easier to just make it a full box, not that much additional cost (just the muck away really as the concrete would have more or less been the same, just different geometry.

 

Once we had the extra space (under the deck) it made sense to utilise it for rainwater harvesting, so the plan was born. The in ground tank systems are probably more robust but do come at a price.

[Nelliekins]

OK, so I have made a start on putting this all together... 

 

 

The PEX pipe adjacent the TMV is a placeholder until I get more check valves, at which time it'll be replaced with copper. 

 

Still waiting for flow switch to be delivered, and connections made to boiler which lives in the room below... 

 

The series of holes in the floor are for the pipes to go up to the hot water manifold, which will be supplied from the TMV. 

 

So, a couple of questions :

 

1. Should I use a Y strainer in the cold input? I had thought about a magnaclean as an option too, but not sure I would see much if any benefit from it? 

 

2. Any ideas on placement of (and recommendations for make/model of) PRV? 

 

3. Am I right in thinking that the F&E tank needs to be 1m+ above the top of the cylinder, and needs to have at least 5% of the volume of the system it is connected to? Realistically that's about 8L I think. 

 

4. The boiler doesn't appear to have an internal pump, so I have bought a 3rd Wilo Pico pump (one for DHW, one for UFH and one for Boiler)... Sound OK? 

 

5. For the UFH side of things, am I going with a set of 3 position valves (one per manifold)? Where do I position the pump and non return valves so that I can bypass the coil and still circulate water to distribute solar gain?

 

 

Edited January 6, 2019 by Nelliekins
Clarifications

[Nelliekins]

On 03/01/2019 at 09:20, Bitpipe said:

 

IBCs are roughly 1m2 so you could squeeze a few in - we put a mezzanine deck over the top of the IBCs which makes a nice feature.

 

We have 6m of bifold doors that open out straight into the space above our lightwell so a mezzanine deck wasn't really an option. 

 

My plan is to stack the IBCs 2 high to get 4 in there, and it'll all be under the deck... The ladder will go up to an access hatch which will be gas strut supported. If I could make it a walk out area like yours I would do that in a heartbeat, because it'll improve light levels enormously in the basement... 

 

Quote

TBH we originally envisaged our lightwell / egress area as a tiered construction but the SE pointed out that it would be easier to just make it a full box, not that much additional cost (just the muck away really as the concrete would have more or less been the same, just different geometry.

 

Once we had the extra space (under the deck) it made sense to utilise it for rainwater harvesting, so the plan was born. The in ground tank systems are probably more robust but do come at a price.

 

Yeah it was the cost of tanks that pushed me towards IBCs... we are in full - tilt "save every penny we can" mode now... If we have to get another lump of cash from the mortgage provider we won't be able to afford to live there! Hence the economy drive, the homebrew heating system and IBCs.

 

Also bought a set of adapters to utilise our excess UFH pipe (we have about 600m left over, so am running 16mm PEX everywhere, for all the hot, cold and RWH pipes inside I can - fittings are 6 quid a pair and pipe is essentially free now it's paid for!). I find it interesting that every sole that has provided their own quantities for supply have overestimated by a typical 20% (be it the floorboards, or the ICF, or the UFH,...) and the stuff I have quantified has always been nearly exactly right... 

 

Edited January 6, 2019 by Nelliekins
Auto correct corrections

[Nelliekins]

On 31/12/2018 at 14:01, PeterW said:

I’d mount the stat about 6” above the bottom inlet directly opposite the outlet. It will probably be mid point of the coils for the UFH at that point too. 

 

 

@PeterW I had a thought about mounting a number of DS18B20 temp sensors in the system in tee (actually branch F) fittings, which might help with fine tuning of the system temps and flow rates. Specifically:

 

- 1 in each of boiler flow & return connections to cylinder

- 1 in each of top and bottom connections for coil

 

Is there anywhere else that would be of use to me? 

 

I would still have at least 1 cylinder stat (probably 2, with the second at same height as, or just above, the boiler connection at the top, but opposing) so that I don't need the automaton system to keep hot water and heating working! 

 

Does this make sense? 

Edited January 6, 2019 by Nelliekins
Clarifications

[Nelliekins]

Right, I have finally arranged for a plumber to come and connect up the boiler. As of Weds, we should have the boiler connected to both the gas and the cyli^H^H^H heat bank. ?

 

I have a F&E tank ready to plumb in above the heat bank, and a manifold ready to connect to the hot water outlet (although that's not needed prior to the boiler being commissioned). I've got a single cylinder stat (a Drayton HTS3) to be positioned approx 6" above the bottom inlet on opposite side of heat bank (as per suggestion by @PeterW), although am still considering having a second stat... I don't seem to be able to find any information on the hysteresis for the stat, but presumably I want it to trigger a demand for heat at around 50C (when the PHE can no longer provide DHW at a high enough temperature), and to tell the boiler it is satisfied at around 65-70C? Do I need 2 stats to achieve this?

 

Once the hot water is working, I need to progress the UFH side of things, which will be done using the coil in the bottom of the heat bank. I'm thinking of 2 pumps - one for the basement+ground floor, and one for the first floor. That separates the 2 types of emitters, and pretty much balances the length of pipe being serviced by each pump (because the basement offsets the integral garage area which isn't heated).

 

So, from the coil, I am thinking we want to have a tee, so that the 2 separate "zones" are parallel. From each leg of the tee, there will be a motorised zone valve followed by a pump, followed by the manifold.

 

Questions:

1. Will the 2 pumps be sufficient? We're only looking at a relatively low flow rate, with a low flow temperature (25-28C), but it's a lot of pipe to pump through... We've got Wilo Pico glandless pumps following the earlier recommendation
2. If we close both zone valves, we'll have effectively sealed off the coil - I presume that since the coil will still be getting hot, we will need a small expansion vessel and a PRV - is that correct?
3. Do we need an expansion vessel for the rest of the UFH circuit as well? IIRC we are intending to run the UFH at 1 bar.
4. I presume that since there is a drain cock fitted to every manifold, we don't need to add additional drainage points for the UFH - is that correct?
5. On the subject of drain cocks, I presume we will need one on the DHW side of things, so we can drain down the heat bank?

Anything else I need to know / do?

[PeterW]

Right...

 

Questions:

1. Will the 2 pumps be sufficient? We're only looking at a relatively low flow rate, with a low flow temperature (25-28C), but it's a lot of pipe to pump through... We've got Wilo Pico glandless pumps following the earlier recommendation - yes, they will be adequate although you may want to keep the speed evenly balanced. How will you create the low flow temp though..?
2. If we close both zone valves, we'll have effectively sealed off the coil - I presume that since the coil will still be getting hot, we will need a small expansion vessel and a PRV - is that correct? Sort of but you only zone the flow not the return so there is always expansion there - you need to calculate the volume properly as it’s the whole of the UFH and rads not just the coil. 
3. Do we need an expansion vessel for the rest of the UFH circuit as well? IIRC we are intending to run the UFH at 1 bar. See above - doubt you will find a PRV at less than 1.5bar and it may just be quicker and easier to use the F&E assuming its higher than the top of the radiators ..??
4. I presume that since there is a drain cock fitted to every manifold, we don't need to add additional drainage points for the UFH - is that correct? Yes although draining UFH is nigh on impossible anyway so don’t panic. 
5. On the subject of drain cocks, I presume we will need one on the DHW side of things, so we can drain down the heat bank? Yes

 

 

[Nelliekins]

Just now, PeterW said:

Right...

 

Questions:

1. Will the 2 pumps be sufficient? We're only looking at a relatively low flow rate, with a low flow temperature (25-28C), but it's a lot of pipe to pump through... We've got Wilo Pico glandless pumps following the earlier recommendation - yes, they will be adequate although you may want to keep the speed evenly balanced. How will you create the low flow temp though..?

The coil in the cylinder will heat up to the temperature of the stored primary water. I was kind of assuming that by modulating the flow (or changing the rate through the coil) I would be able to control the amount of heat that is introduced to the UFH. Does that make sense, or am I being naive? We want to run the floor at around 25C according to @JSHarris spreadsheet. We could always use a mixing valve to keep the temp down I guess? 

1. If we close both zone valves, we'll have effectively sealed off the coil - I presume that since the coil will still be getting hot, we will need a small expansion vessel and a PRV - is that correct? Sort of but you only zone the flow not the return so there is always expansion there - you need to calculate the volume properly as it’s the whole of the UFH and rads not just the coil. 

I think we worked out that there was approx 250 litres of water in the UFH system. (Pex pipe is 16mm OD, 12mm ID, and 2.2km installed gives 0.249m3). IIRC we need to support 5% for expansion (12.5 litres) so is an 18l expansion vessel appropriate? 

 

I had assumed that we would have a 3 port diverter for each UFH circuit, with it switching between coil flow and return, so that we can circulate the water without heating it. But that would not work for expansion, would it? 

1. Do we need an expansion vessel for the rest of the UFH circuit as well? IIRC we are intending to run the UFH at 1 bar. See above - doubt you will find a PRV at less than 1.5bar and it may just be quicker and easier to use the F&E assuming its higher than the top of the radiators ..??

No rads. We are planning on running 3 manifolds of UFH off the coil in the DIY heat bank. Our peak space heating load is 3.6kw @ 25C differential, but accounting for us, the cats and the MVHR it's more like 2.5kw needed from the coil (sound OK?). The F&E we have now is for the DHW side only, and comes off the DIY heat bank. 

1. I presume that since there is a drain cock fitted to every manifold, we don't need to add additional drainage points for the UFH - is that correct? Yes although draining UFH is nigh on impossible anyway so don’t panic. 
2. On the subject of drain cocks, I presume we will need one on the DHW side of things, so we can drain down the heat bank? Yes

OK done. It's been put on a tee at the bottom of the DIY heat bank. 

 

 

[Nelliekins]

Just now, Nelliekins said:

Will the 2 pumps be sufficient? We're only looking at a relatively low flow rate, with a low flow temperature (25-28C), but it's a lot of pipe to pump through... We've got Wilo Pico glandless pumps following the earlier recommendation - yes, they will be adequate although you may want to keep the speed evenly balanced. How will you create the low flow temp though..?

 

Would having just a single pump (or 2 in parallel) work better than 2 separate pumped UFH circuits in this setup? That way we would have a single 3 port diverter valve, with the 3 ports being

 

1. Flow from coil in DIY heat bank

2. Flow to UFH

3. Return from UFH to coil (teed off?) 

 

The pump and the EV would be downstream of the diverter valve. This would allow the EV to stay in circuit no matter what position the valve was in, and also allow us to pump water without heating it (eg to spread heat from solar gain). 

[PeterW]

You need a mixing or blending valve not a 3 way here as otherwise you will get whatever temperature the tank gives you and that’s not a good idea. Modulating pump speed to make the temperature differential won’t work. 

 

Pump per manifold circuit

Single blender on the output from the tank coil

zone valve per zone. 

 

Easy to set up and should work fine. 

[Nelliekins]

On 26/01/2019 at 09:40, PeterW said:

You need a mixing or blending valve not a 3 way here as otherwise you will get whatever temperature the tank gives you and that’s not a good idea. Modulating pump speed to make the temperature differential won’t work. 

 

Pump per manifold circuit

Single blender on the output from the tank coil

zone valve per zone. 

 

Easy to set up and should work fine. 

 

@PeterW ok I have fitted 2-port valves to each of the circuits, and pumps inline with each valve (driven by the valves). Here's a photo :

 

(ignore the wiring and plumbing chaos - there's a rainwater distribution manifold and 2 wiring centres yet to fit!) 

 

So, what we have is coil going to tee on both flow and return. One loop formed by bypass valve set at 2 bar (little black thing by cylinder stat) . Other loop formed by Reliance mixing valve rated for 25-50C output, which then gets split to go into the circuits for UFH pumping. 

 

It all seems to work, except for one detail... 

After 30 minutes of running, the flow on the manifolds was 20C and return 12C. Fine, but after 24 hours of running the flow was still 20C with return up to around 17C.

 

Clearly 20C isn't quite warm enough for the UFH - the spreadsheet from @JSHarris said 25C was the goal for our house. With one pump knocked off (thereby getting all the heat from the coil into just one half of the floor) it still stayed at 20C.

 

My guess is that I have a few options:

 

1. Increase the cylinder temp from 55C to around 70C to see if that gets more heat into the coil

2. Impede the flow rate somehow so that the water travels through the coil slower (pumps already at minimum rate though!) 

3. Replace the cylinder with one that has a more useful coil

 

Have I missed anything? Going to put some DS18B20s into the thermowells all over the system tonight to see if I am missing anything simple... 

[PeterW]

If that is a standard hot water tank then the coil won’t be big enough to give heat to the UFH loop at 50c.  I would up it to 70c and see how the UFH reacts. 

[ProDave]

That coil looks to be at the bottom suggesting it is a heat input coil in an ordinary indirect tank.  A heat output coil in a thermal store would be at the top of the tank.

[PeterW]

5 minutes ago, ProDave said:

That coil looks to be at the bottom suggesting it is a heat input coil in an ordinary indirect tank.  A heat output coil in a thermal store would be at the top of the tank.

 

Not necessarily

 

coil at the top is normally for DHW so needs to be in the 60c zone. UFH can be lower but that tank won’t hold that much heat capacity so needs to be hotter. 

[Nelliekins]

10 hours ago, ProDave said:

That coil looks to be at the bottom suggesting it is a heat input coil in an ordinary indirect tank.  A heat output coil in a thermal store would be at the top of the tank.

 

It is indeed an ordinary indirect cylinder, 140L stainless steel RM Aquastel. 

 

10 hours ago, PeterW said:

 

Not necessarily

 

coil at the top is normally for DHW so needs to be in the 60c zone. UFH can be lower but that tank won’t hold that much heat capacity so needs to be hotter. 

 

Jeremy's spreadsheet said 25-26°C floor temp was what we were aiming for. Mixing valve is good for 25-50°C, and was set at 3/10. Flow was steady 20°C and return around 17°C.

 

I boosted the cylinder temp from 55°C to 65°C by turning up the wick a bit on Mr Boiler, and then went to look at the mixing valve...  I set it midway (5/10) at 5:30pm today. By 7pm, flow was up from 20°C to 28°C and return was 22°C on concrete floors and 26°C on 1st floor. 

 

So fiddling with the mixing valve and giving the boiler a bit more gas seems to have sorted the UFH temps... ?

 

Still suspicious that it won't cut the mustard if it gets properly cold outside, but for now we have heating! 

 

[Nelliekins]

@PeterW I am beginning to wonder if the 2 port valve were the best idea...

 

How will we circulate the water without taking in heat from the cylinder, e.g. to distribute solar gain from the kitchen? 

 

Am thinking that a 3 way diverter (not mid-position) valve would either connect a circuit to the flow from the coil or "short across" to the return. 

 

Thinking about it, we would need a 2 port valve as well as the 2no. 3-port valves, to close off the return that goes to the coil, correct? 

[dpmiller]

9 hours ago, Nelliekins said:

@PeterW I am beginning to wonder if the 2 port valve were the best idea...

 

How will we circulate the water without taking in heat from the cylinder, e.g. to distribute solar gain from the kitchen? 

 

Am thinking that a 3 way diverter (not mid-position) valve would either connect a circuit to the flow from the coil or "short across" to the return. 

 

Thinking about it, we would need a 2 port valve as well as the 2no. 3-port valves, to close off the return that goes to the coil, correct? 

What about using a bypass valve (tee'd between pump and motor valve) to open if the pump's running with the 2-port closed? That would give you circulation without flow through the coil.

[Nelliekins]

1 hour ago, dpmiller said:

What about using a bypass valve (tee'd between pump and motor valve) to open if thepump's running with the 2-port closed? That would give you circulation without flow through the coil.

 

Hmmm... That might work you know! Good idea! Just have to work out how to fit 2 into the same part of the return...  ?