Re-designing my heating system

[SimonD]

36 minutes ago, John Carroll said:

The boiler dT of 9.5C is because

 

Yes, I know the formulas to calculate the mass flow rates etc, but maybe my question and empirical experience wasn't clear enough.

 

39 minutes ago, John Carroll said:

but the boiler shouldn't cycle

 

This is really the crux of it. There is the assumption that the boiler only cycles when, as you say, minimum output is greater than the load. However, my empirical observations suggest that something else goes on depending on boiler manufacturer implementation boiler modulation and controls. And this means that at low flow temperatures, the boiler sees a return temperature it thinks is too high relative to the also low flow temperature and therefore short cycles, even when the load is greater than the minimum boiler output. This is pretty much what I have observed and I don't quite know why yet. At the end of the day we're talking about a dynamic system that rarely works steady state, so return temp often fluctuates. Therefore, if the return temp is so close to flow temperature, there is less margin for fluctuations. Again I've seen this happen fairly often in the real world where a gas boiler is cycling and the internal return temperature sensor thinks it is close or equal to flow temperature. However, if I connect my sensors I can see a good 10C differential - this obviously happens when we get close to output/load limits or flow rates are too high through the system.

 

So my essential question is whether it is appropriate to design for a dT that is less than 1/2 that recommended by most boiler manufacturers and CIBSE for example and does this actually work reliably and efficiently in practise?

[JohnMo]

41 minutes ago, SimonD said:

means that at low flow temperatures, the boiler sees a return temperature it thinks is too high relative to the also low flow temperature and therefore short cycles

 

I know when I was running a boiler last year (Atag A325ECX), the flow temp was set to 36 degs, the pump speed was being modulated by the boiler and it set the dT at about 6 degs. This also coincides with the part load performance data on the datasheet (30/36). But also think min modulation kW allowed is also considerably higher than the 50/30 flow temps. Think it must try to keep the return temp at 30 and above, but just reduces the dT to accommodate a lower flow temp set point.

[John Carroll]

I know Vaillants have a internal bypass that defaults to 2.5M and is adjustable to I think a max of 3.5M, some also have a HEX dP of 4.0M at max flow rate, this is why 20C is used as the dT, a 20kw boiler will have a max flowrate of, (20*860/60/20), 14.3LPM, a 32kw boiler, 23LPM etc, most boilers now have a 7M pump and will still have a head of ~ 6.5M at the rated boiler flowrate, this then leaves around 2.5/3.0M to circulate the water through the system, more than that or even less, in some systems, a LLH or CCTs will then have to be installed, the primary side will deal with the boiler HEX dP and another circ pump on the secondary side looks after the system losses. The dP through a HEX designed for a 4M dP at say 20LPM will only have a dP of 0.56M at 15LPM and only 0.25M at 10LPM, so the bypass will be open for a lot of the time, 12LPM will result in a ~ 10kw rad output but because the HEX dP is only 0.25M then there is around 5.5M/6M remaining which will cause quite a lot of recirc with very high return temps at the boiler so you can be measuring a dT of 10C at the boiler flow/return but the actual dT might only be a few degrees so maybe this why the boiler trips, its easy to check that out where you can read off the return (as well as the flow) temperature in the boiler menu and compare them with the measured temperatures.

 

The newer vaillants now have a settable dT of between 10C & 20C which modulates the pump to a minimum speed of 50%. I know someone with this boiler but cannot achieve this, he is now trying the pump on manual set to 50% so awaiting results.

Edited March 7 by John Carroll

[John Carroll]

6 hours ago, JohnMo said:

 

I know when I was running a boiler last year (Atag A325ECX), the flow temp was set to 36 degs, the pump speed was being modulated by the boiler and it set the dT at about 6 degs. This also coincides with the part load performance data on the datasheet (30/36). But also think min modulation kW allowed is also considerably higher than the 50/30 flow temps. Think it must try to keep the return temp at 30 and above, but just reduces the dT to accommodate a lower flow temp set point.

This is very interesting but I suppose there should really be a new thread.

 

Even if the circ pump is not controlled/modulated then the  dT will decrease with reduced flow temperature.

If you use a fixed flow rate of 1.0LPM/kw then WC will result in the following example temperatures,  flow/return/dT...... 65.0C/54.4C/10.6C.

45.0C/39.9C/5.1C. 40.0C/36.2C/3.8C. (Assumes a required room temp of 20C)

Edited March 7 by John Carroll

[jayc89]

23 hours ago, JoeBano said:

Have a look at this video, i could be wrong but looks very similar what you have now.

 

 

That looks very similar to the conclusions I've started to draw here. I assume the "Kimbo Pipe", is the oversized CCT at the boiler arrangement already discussed in here? A large piece of that design seems to be able to have different flow rates for DWH, Rads AND UFH, whereas, AFAIK, my boiler is only capable of two different flow rates; max when DWH is calling for heat and a weather curve otherwise. 

 

The video also suggests that the electronic blending valve on the UFH flow is also on a weather curve, controlled by the boiler. It was hard enough to find a weather comp unit for my boiler (Baxi discontinued it a couple of years ago), I highly doubt it will natively support a proprietary blending valve too. Looking at the ESBE actuators, it looks like you can fit a separate, external, weather comp sensor to it, but I then worry two weather comp units run the risk of working against each other, different curves etc. 

 

In the video it's also suggested to do away with 2 port zone valves, but surely something's still required on the rad circuit to avoid it receiving heat when only the UFH is calling for heat, no? 

 

19 hours ago, John Carroll said:

 

This may (or may not) be of some help of seeing the "total" picture, I configured this when my daughter was installing UFH and just tailored it to your UFH + rad heating demands, I'm not sure when you say spec'd for DT25 if you mean rads designed as "T25" rads?, if so they will give a rad output of 40.6% of T50 sized rads and would require a oversizing factor of X2.46. say 2.5. I'm also assuming that the required boiler flow temperature required is 48C and the UFH required manifold temp is 40C. I didn't include cylinder heating.

 

 

 

 

That looks much better than my rudimentary calcs, thanks for sharing. Interesting that the max flow rate of the system wouldn't be much more than the boiler pump should be able to supply. I assume it'd still be best to keep the separate UFH pump as to not have the boiler pump running at max when both rads and UFH are calling for heat. 

[JohnMo]

47 minutes ago, jayc89 said:

video also suggests that the electronic blending valve on the UFH flow is also on a weather curve,

I have messed with WC on UFH for 2 years, the truth is you really don't need it. A good mixer valve like an IVAR is all you need. Have WC for the rads. I have a standalone ESBE mixer and controller they are good, but stupid expensive. Mine is there for a specific reason, not normally needed.

 

Then as suggested above normally closed 2 port valve for the cylinder and normal open for the heating system, when there is a call for DHW heating the 2P valve opens and the others shut. Boiler to high output. Not sure you need to do anything else. Anymore looks to be for zero gain. 

 

Normally open 2P should just be an actuator change.

 

 

[SimonD]

16 hours ago, John Carroll said:

I know Vaillants have a internal bypass that defaults to 2.5M and is adjustable to I think a max of 3.5M, some also have a HEX dP of 4.0M at max flow rate, this is why 20C is used as the dT, a 20kw boiler will have a max flowrate of, (20*860/60/20), 14.3LPM, a 32kw boiler, 23LPM etc, most boilers now have a 7M pump and will still have a head of ~ 6.5M at the rated boiler flowrate, this then leaves around 2.5/3.0M to circulate the water through the system, more than that or even less, in some systems, a LLH or CCTs will then have to be installed, the primary side will deal with the boiler HEX dP and another circ pump on the secondary side looks after the system losses. The dP through a HEX designed for a 4M dP at say 20LPM will only have a dP of 0.56M at 15LPM and only 0.25M at 10LPM, so the bypass will be open for a lot of the time, 12LPM will result in a ~ 10kw rad output but because the HEX dP is only 0.25M then there is around 5.5M/6M remaining which will cause quite a lot of recirc with very high return temps at the boiler so you can be measuring a dT of 10C at the boiler flow/return but the actual dT might only be a few degrees so maybe this why the boiler trips, its easy to check that out where you can read off the return (as well as the flow) temperature in the boiler menu and compare them with the measured temperatures.

 

 

This does highlight a big issue in that design approaches often don't cover the role that HEX dP and bypass arrangements within the boiler play in selecting for hydraulic separation or not, and that it's more than just down to flow rate differential between heating circuits. I had exactly this problem in a heating system I had to rectify a couple of months back where they'd had a good half dozen plumbers round to try and fix it without joy. I suspected that a major problem was HEX dP. I tried to get the pressure loss data from the boiler manufacturer's technical department who said they'd email it to me and eventually received a pump chart for a system boiler and not the HEX data for the boiler I was dealing with! No HEX data could be extracted from them. In the end I was proven to be correct and got the system working properly for the first time ever.

 

But as I suggested earlier, it is also a warning that we should not assume that we can merrily design and commission gas boiler systems to run at less than 1/2 specified boiler dT 20.

 

I always use my own measuring equipment in parallel to the boiler's to get a proper picture of what is going on.

 

17 hours ago, JohnMo said:

I know when I was running a boiler last year (Atag A325ECX), the flow temp was set to 36 degs, the pump speed was being modulated by the boiler and it set the dT at about 6 degs. This also coincides with the part load performance data on the datasheet (30/36). But also think min modulation kW allowed is also considerably higher than the 50/30 flow temps. Think it must try to keep the return temp at 30 and above, but just reduces the dT to accommodate a lower flow temp set point.

 

Atags are interesting things.I first came across a behemoth of one a while back which is a model sold by the 'other' Atag company in the UK which is more focussed on the commercial boilers. It was an amazing piece of machinery which did some pretty clever things with its built in modulation and that on start up it modulates from minimum output and ramps up from there as the system warms up. At some point I might get trained on their boilers to understand them better, particularly since they now provide an 18 year warranty.

 

[John Carroll]

1 hour ago, SimonD said:

 

This does highlight a big issue in that design approaches often don't cover the role that HEX dP and bypass arrangements within the boiler play in selecting for hydraulic separation or not, and that it's more than just down to flow rate differential between heating circuits. I had exactly this problem in a heating system I had to rectify a couple of months back where they'd had a good half dozen plumbers round to try and fix it without joy. I suspected that a major problem was HEX dP. I tried to get the pressure loss data from the boiler manufacturer's technical department who said they'd email it to me and eventually received a pump chart for a system boiler and not the HEX data for the boiler I was dealing with! No HEX data could be extracted from them. In the end I was proven to be correct and got the system working properly for the first time ever.

 

But as I suggested earlier, it is also a warning that we should not assume that we can merrily design and commission gas boiler systems to run at less than 1/2 specified boiler dT 20.

 

I always use my own measuring equipment in parallel to the boiler's to get a proper picture of what is going on.

 

 

Atags are interesting things.I first came across a behemoth of one a while back which is a model sold by the 'other' Atag company in the UK which is more focussed on the commercial boilers. It was an amazing piece of machinery which did some pretty clever things with its built in modulation and that on start up it modulates from minimum output and ramps up from there as the system warms up. At some point I might get trained on their boilers to understand them better, particularly since they now provide an 18 year warranty.

 

What make/model and minimum output is your boiler?.

[SimonD]

2 hours ago, John Carroll said:

What make/model and minimum output is your boiler?.

 

I'm not referring to my boiler, just talking about general principles of system design as I design and install systems myself (although I'm fairly new to this business). But it seems I'm recently in more demand to fix existing installations that don't work properly, especially newish builds with stupidly large boilers installed with no modulation and a random mix of poorly design UFH and rads, and nobody else wants to touch them.

[jayc89]

13 minutes ago, SimonD said:

 

I'm not referring to my boiler, just talking about general principles of system design as I design and install systems myself (although I'm fairly new to this business). But it seems I'm recently in more demand to fix existing installations that don't work properly, especially newish builds with stupidly large boilers installed with no modulation and a random mix of poorly design UFH and rads, and nobody else wants to touch them.

 

That's pretty much how I've ended up where I am.

 

Before we knew better we got a local plumber to fit a new system; boiler, UVC, rads. All spec'd to, what I think was, a 70c flow temp. Even at that, all bedroom rads were undersized and have been replaced as part of going down to DT25. Whilst he was commissioning the boiler/tank we found the ground floor joists were rotten so had to deal with that and decided on a slab + UFH as a replacement. He didn't want to know anything about the UFH and just offered to leave a terminated flow/return pipe for someone else to connect into. This is why I've always been concerned the system hasn't been running as well as it could be. 

 

There are a lot of old school trades out there that don't necessarily do a bad job (although I'm far from pleased in my case), but they certainly don't work to more modern practices. 

[SimonD]

1 hour ago, jayc89 said:

 

That's pretty much how I've ended up where I am.

 

Before we knew better we got a local plumber to fit a new system; boiler, UVC, rads. All spec'd to, what I think was, a 70c flow temp. Even at that, all bedroom rads were undersized and have been replaced as part of going down to DT25. Whilst he was commissioning the boiler/tank we found the ground floor joists were rotten so had to deal with that and decided on a slab + UFH as a replacement. He didn't want to know anything about the UFH and just offered to leave a terminated flow/return pipe for someone else to connect into. This is why I've always been concerned the system hasn't been running as well as it could be. 

 

There are a lot of old school trades out there that don't necessarily do a bad job (although I'm far from pleased in my case), but they certainly don't work to more modern practices. 

 

Yup, we've trodden a similar path and this is one reason I now do what I do. I couldn't find anyone locally I trusted to actually design my new system for my build, so I decided to take it on myself. While I was doing so, a friend who has a heating business but a long way from me suggested I do the training and get my tickets, so I did. I only do it part time because I'm still building. It's a steep learning curve and much research needed in some cases but I prefer working on complex problems over chucking in replacement boilers as I find it more stimulating and satisfying.

[John Carroll]

52 minutes ago, SimonD said:

3 hours ago, SimonD said:

 

I'm not referring to my boiler, just talking about general principles of system design as I design and install systems myself (although I'm fairly new to this business). But it seems I'm recently in more demand to fix existing installations that don't work properly, especially newish builds with stupidly large boilers installed with no modulation and a random mix of poorly design UFH and rads, and nobody else wants to touch them.

 

A huge proportion of UK gas boilers are combi so one would think that all these would be, in the main, very oversized, a fairly modest 32kw  might only have a 15kw heating load, most can be range rated but if cycling, for whatever reason, will still fire up at ~ 65% of their  rated output which will, more than likely cause problems if WC and only looking for say 35C/40C flow temperature.

[jayc89]

Is there a DT that's too low for a gas boiler? If I understand correctly, to get full efficiency from a condensing boiler, you want the flow above the dew point, but the return as cold as possible, so quite a wide DT. I.e. flow 55 and DT 20, for a return of 35? But I've also heard suggestions of some manufacturers now moving towards DT15 which contradicts this. 

 

Ideally, I'd have the system set up to run the same flow rate for both my rads and UFH, removing the UFH mixer (perhaps a pipe stat and zone valve as a fail safe to protect my tiles), and letting weather comp do it's thing, but to achieve this, assuming my calcs are correct, I'd need a boiler flow of 45 and a DT of 10, with a radiator DT of 19 (which makes them pretty big!)

 

Whilst it simplifies the system, allowing me to remove mixing valves etc, am I sacrificing efficiency? 

Edited March 11 by jayc89

[JohnMo]

57 minutes ago, jayc89 said:

to get full efficiency from a condensing boiler, you want the flow above the dew point

No it's based on the return temperature not the flow temperature 

[jayc89]

2 minutes ago, JohnMo said:

No it's based on the return temperature not the flow temperature 

 

My assumption was based on this - https://www.heatgeek.com/condensing-boilers-efficiency/

 

Quote

However, there is a caveat! When we get condensing occurring the humidity drops, this, in turn, lowers our dew point (we want it as high as possible to extract more latent heat) and slows or halts the condensing process. So, if for example, we had a very long heat exchanger, a dewpoint of 55°c and the temperature of the heat exchanger is an even 54°c all over, you may suppose we could eventually condense all the vapour into the water if the heat exchanger was long enough. In actual fact as soon as the humidity drops slightly the dew point decreases and condensing will theoretically halt for the rest of the heat exchange as the air is no longer 'saturated' with H20. Hence it is below the relative point of saturation.
 

For this reason, we need to target as lower a return temperature as possible to maximise the condensing effect, not the flow temperature. The way we achieve this is by utilising ERP modulating pumps to maintain a nice wide delta T 20 across the flow and return to the boiler. In other words, a nice slow flow rate to make sure the return is as cool as possible.

 

Suggesting a "wide" DT is required, and a DT10 would be more inefficient than a DT of 20? 

[jayc89]

Separate question, I have the upstairs heat loss, and applied a correction factor to it to work out the radiator sizes required at DT19 (well, I actually used DT20, for simplicity).

 

 

When spec'ing the flow/return pipe sizes, which set of numbers should I be using? The actual heat loss numbers suggest 22mm teeing off in 15mm to the radiators will be fine, whereas the "corrected" numbers suggest I'd need 28mm from the boiler with an intermediate branch of 22mm (to supply the master suite), before teeing off to 15mm for individual radiators. 

[SimonD]

3 hours ago, jayc89 said:

Separate question, I have the upstairs heat loss, and applied a correction factor to it to work out the radiator sizes required at DT19 (well, I actually used DT20, for simplicity).

 

 

When spec'ing the flow/return pipe sizes, which set of numbers should I be using? The actual heat loss numbers suggest 22mm teeing off in 15mm to the radiators will be fine, whereas the "corrected" numbers suggest I'd need 28mm from the boiler with an intermediate branch of 22mm (to supply the master suite), before teeing off to 15mm for individual radiators. 

 

You need to use the actual calculated heat output of the radiator/s for each branch and circuit. I'm assuming that your correction factor is the catalogue radiator output given at dt50?

[jayc89]

9 hours ago, SimonD said:

 

You need to use the actual calculated heat output of the radiator/s for each branch and circuit. I'm assuming that your correction factor is the catalogue radiator output given at dt50?

 

Yeah exactly that, which gives me a total max heat transfer of 69,000 BTU at dT50 and a max heat transfer on the Master Suite branch (bedroom, wardrobe and 2x towel radiators) of 25,000BTU at dT50. 

 

From what I've seen online copper is good for;

- 15mm up to 13,000 BTU

- 22mm up to 36,000 BTU

- 28mm up to 64,000 BTU

 

So the majority of runs are clear; 22mm for the Master Suite branch, 15mm tees for all single emitters (which we already have) albeit Bedroom 1 and 2 are pushing the max capacity of a 15mm run, but the main "trunk" from the boiler (which only has 22mm tappings) looks like it could be a struggle. Do I really need to run that in 35mm? 

[John Carroll]

16 hours ago, jayc89 said:

Is there a DT that's too low for a gas boiler? If I understand correctly, to get full efficiency from a condensing boiler, you want the flow above the dew point, but the return as cold as possible, so quite a wide DT. I.e. flow 55 and DT 20, for a return of 35? But I've also heard suggestions of some manufacturers now moving towards DT15 which contradicts this. 

 

Ideally, I'd have the system set up to run the same flow rate for both my rads and UFH, removing the UFH mixer (perhaps a pipe stat and zone valve as a fail safe to protect my tiles), and letting weather comp do it's thing, but to achieve this, assuming my calcs are correct, I'd need a boiler flow of 45 and a DT of 10, with a radiator DT of 19 (which makes them pretty big!)

 

Whilst it simplifies the system, allowing me to remove mixing valves etc, am I sacrificing efficiency? 

Your total heating demand is 16kw with a required flowrate of 23 LPM.

All boiler makers base their max flowrate based on a dT of 20C, not because they care about condensing but because the boiler HEX head loss is based on this, a typical max HEX loss might be 3.5M at max flow. If the installed boiler is rated at 20kw then its max flowrate is, 20*860/60/20, 14.3LPM, if the HEX dP is 3.5M at this flowrate, then its 3.5*(23/14.3)^2, 9.1M at a flowrate of 23LPM, a non runner maybe even with a LLH, this assumes you require both CH&UFH on together. (if my calcs are correct).

 

Edited March 12 by John Carroll

[jayc89]

10 minutes ago, John Carroll said:

Your total heating demand is 16kw with a required flowrate of 23 LPM.

All boiler makers base their max flowrate based on a dT of 20C, not because they care about condensing but because the boiler HEX head loss is based on this, a typical max HEX loss might be 3.5M at max flow. If the installed boiler is rated at 20kw then its max flowrate is, 20*860/60/20, 14.3LPM, if the HEX dP is 3.5M at this flowrate, then its 3.5*(23/14.3)^2, 9.1M at a flowrate of 23LPM, a non runner even with a LLH? (if my calcs are correct).

 

 

Boiler is 33kw, massively oversized for what we need, but that's another story. I believe it's max heating output is 30.3kw, so using your calcs;

 

30.3*860/60/20 = 21.72 LPM

 

3.5*(23/21.72)^2 = 3.924678395

 

Presumably a LLH, or CCT, on the heating loop would be required, with an additional pump for the rads?  

 

Anything running at max, instinctively, feels like it'd be inefficient though? 

[SimonD]

1 hour ago, jayc89 said:

 

Yeah exactly that, which gives me a total max heat transfer of 69,000 BTU at dT50 and a max heat transfer on the Master Suite branch (bedroom, wardrobe and 2x towel radiators) of 25,000BTU at dT50. 

 

From what I've seen online copper is good for;

- 15mm up to 13,000 BTU

- 22mm up to 36,000 BTU

- 28mm up to 64,000 BTU

 

So the majority of runs are clear; 22mm for the Master Suite branch, 15mm tees for all single emitters (which we already have) albeit Bedroom 1 and 2 are pushing the max capacity of a 15mm run, but the main "trunk" from the boiler (which only has 22mm tappings) looks like it could be a struggle. Do I really need to run that in 35mm? 

 

I don't work in BTUs I'm afraid, I mostly grew up and went to school in Europe and by the time I got back here, we'd moved to metric too. 😉

 

Your pipe sizing is going to be based on the flow velocity through the pipes and resultant pressure loss, not a rule of thumb about how much a pipe can carry. At dT 20 this is going to be based on a flow rate of approx. 11.5 lpm. Presuming you can range rate/modulate your 30kW boiler? If not, then as said, at 30kW it's approx. 21.5 lpm.

 

Now what you need to do is draw up your pipework with lengths and calculate the flow through each section according to what heat load that section is going to carry and then look at the flow velocity according to a pipe sizing chart. Then you can size your pipework for an appropriate size to carry the heating load at a reasonable velocity. Preferrably choose a velocity just below 1m/s. Once you have this figure, you can then look at your pressure loss in the pipework and decide on the pipe diameter.

 

I think that with your UFH, you should ideally go with some form of hydraulic separation in your system.

[JohnMo]

36 minutes ago, SimonD said:

don't work in BTUs I'm afraid

Me neither and I am over 60, soon as saw BTU I moved to the next thread. Leave BTU to USA.

[jayc89]

2 hours ago, jayc89 said:

 

Yeah exactly that, which gives me a total max heat transfer of 20.2kW at dT50 and a max heat transfer on the Master Suite branch (bedroom, wardrobe and 2x towel radiators) of 7.3 kW at dT50. 

 

From what I've seen online copper is good for;

- 15mm up to 3.8kW

- 22mm up to 10.5kW

- 28mm up to 18.7kW

 

So the majority of runs are clear; 22mm for the Master Suite branch, 15mm tees for all single emitters (which we already have) albeit Bedroom 1 and 2 are pushing the max capacity of a 15mm run, but the main "trunk" from the boiler (which only has 22mm tappings) looks like it could be a struggle. Do I really need to run that in 35mm? 

 

Fixed it

Forgiven, @SimonD @JohnMo?
 

[SimonD]

1 hour ago, jayc89 said:

From what I've seen online copper is good for;

- 15mm up to 3.8kW

- 22mm up to 10.5kW

- 28mm up to 18.7kW

 

1 hour ago, jayc89 said:

Fixed it

 

 

Hmmm, those figures are a long way out if you're working on a dT of 20 in your heating system.

 

Assuming full load is equivalent to your boiler at 30kW output, you're looking at a mass flow rate of approx 0.36kg/s. On this basis, using a 22mm copper pipe you'll get over 1m/s velocity, lets say 1.25m/s. This is a bit fast, but not completely out of the question. Best look at 28mm primary pipework as this will give you around 0.8m/s which is almost ideal. Pressure loss for the 28mm pipe will be 0.021 meter head/m of pipe length. At 22mm it would be 0.070 meter head/m. At 35mm pipework, flow rate would be so low, it wouldn't be advisable on dT 20 but you'd need to look at the values for heatpump dT

 

This is about right as some boiler manufacturers suggest using 28mm primary pipework above 26kW boiler output.

 

As you can see, the choice of pipe size is a bit more involved.

 

When I designed my own system I ran the calcs on the basis of dT20, dT5 & dT7 to make sure my pipework would be fine with gas boiler or heatpump, or even a hybrid system.

 

Obviously there will be pretty different figures if you range rate to 16kW

Edited March 12 by SimonD

[jayc89]

42 minutes ago, SimonD said:

 

 

 

Hmmm, those figures are a long way out if you're working on a dT of 20 in your heating system.

 

Assuming full load is equivalent to your boiler at 30kW output, you're looking at a mass flow rate of approx 0.36kg/s. On this basis, using a 22mm copper pipe you'll get over 1m/s velocity, lets say 1.25m/s. This is a bit fast, but not completely out of the question. Best look at 28mm primary pipework as this will give you around 0.8m/s which is almost ideal. Pressure loss for the 28mm pipe will be 0.021 meter head/m of pipe length. At 22mm it would be 0.070 meter head/m. At 35mm pipework, flow rate would be so low, it wouldn't be advisable on dT 20 but you'd need to look at the values for heatpump dT

 

This is about right as some boiler manufacturers suggest using 28mm primary pipework above 26kW boiler output.

 

As you can see, the choice of pipe size is a bit more involved.

 

When I designed my own system I ran the calcs on the basis of dT20, dT5 & dT7 to make sure my pipework would be fine with gas boiler or heatpump, or even a hybrid system.

 

Obviously there will be pretty different figures if you range rate to 16kW

 

Having pulled them off the internet, I assumed they'd be for dT50 and therefore would need some tweaking for a lower dT. 

 

It makes sense to spec the pipework for a future ASHP too, which I assume would optimally run at dT5, using Heat Geek's Cheat Sheet for pipework at keep velocity below 0.9, it suggests the following pipe sizes would be required;

 

 

If I use the same formula for the UFH supply (minus any correction factor for rad sizes), it suggests that also needs a 35mm run !? They seem pretty beefy. 

 

I will of course calculate the run lengths, elbows, tees etc to confirm these.