System boiler modulation with low loss header LLH

[HRP123]

Hi there, thanks for taking the time to read this. I've been trying to diagnose why my gas consumption is so high and been doing various measurements and tests.

With the UFH, my setup is about 10 room zones connected on two separate UF manifolds linked to one UF pump with a return blending valve from the LLH (picture) - i.e. there is no pump on the manifolds, just at the LLH - each room has an actuator opening each zone from the manifold. Total UF area around 160 sqm with UF pipes on average 150mm spacing (some 200 and some 100mm).

In the middle of the night I positioned a camera to see what calls for heat from the boiler at what time. For about two hours in the middle of the night, only the UF pump turned the boiler on, and I believe only 1 or 2 room zones (about 50-80sqm) would have called for heat (I had set the others back on temperature). I think in theory this should be a heat demand of 9kwh. However the gas consumption for that period shows the boiler was going on full 30kw gas burning per hour (I have ranged rated my 40kw boiler down to around 30kw).

My question is if the UF blending temp was turned down to 40/45 degrees, how should the boiler respond when it is set to a flow of 65 degrees? Will it just keep sending out 65 degrees, and because of the low return temp from the UF manifold into the LLH, the LLH will be sending back a low return temperature to the boiler, so the boiler keeps firing fully instead of modulating down to a lower temperature and consuming less gas?

If so, then would any setup with HW, UFH and radiators from one LLH always cause a boiler to fire too much - since the boiler is set to a higher to satisfy the HW but too high for the UFH and radiators maybe, but cannot modulate down because of the LLH distortion?

Thanks for your help... sorry this may be a rookie question!

 

[ShaunCorbs]

Yes spot on tbh your better off running the ufh wide open and maybe look at sizing your rads for the same temp as the ufh then

 

[HRP123]

Thanks for your opinion! I get what you are saying - goes against mainstream advice of going with low flow temperatures for longer.

Basically I need to find the lowest flow temp that will work for the HW cylinders, and push the max temp possible to the UFH (tiles temp maybe 30 degrees). Then set a long boiler anti cycle time (lets say 10-15 mins) and perhaps a 20 degree flow return anti cycle function (vs default 10 difference on WB) on the boiler so basically cycles less when the return starts rising rapidly, giving some time for the heat to flow through into the UFH and rads?

Cheers

 

[ShaunCorbs]

Thanks for your opinion! I get what you are saying - goes against mainstream advice of going with low flow temperatures for longer.

Basically I need to find the lowest flow temp that will work for the HW cylinders, and push the max temp possible to the UFH (tiles temp maybe 30 degrees). Then set a long boiler anti cycle time (lets say 10-15 mins) and perhaps a 20 degree flow return anti cycle function (vs default 10 difference on WB) on the boiler so basically cycles less when the return starts rising rapidly, giving some time for the heat to flow through into the UFH and rads?

Cheers

Depending on your boiler you could go down the option of having a diverter in the boiler so you can run two temperatures

 

[HRP123]

Depending on your boiler you could go down the option of having a diverter in the boiler so you can run two temperatures

It is a 10 year old 40cdi WB, not sure I can install this feature

 

[gmartine]

Do you need a flow temp of 65 or is that just for (2 x 200l) the HW cylinders.

 

[gmartine]

Do you need a flow temp of 65 or is that just for (2 x 200l) the HW cylinders?

 

[HRP123]

Do you need a flow temp of 65 or is that just for (2 x 200l) the HW cylinders?

Yes the flow temp is for the HW cylinders which I have set to about 50-55 degrees - also have a secondary hot water circulator which keeps turning the boiler on for short period throughout the day (every couple of hours).

 

[John.g]

Your system will still only require 9kw, assuming a manifold dT of 8C then the UFH flowrate will be 16.13LPM at 45C, return 38C with 4.61LPM returning to the boiler giving boiler flow/return flows of 4.61LPM and flow/return temperatures of 65C/38C, 9kw in both but the boilr dT at 27C is getting close to the 30C which normally trips the burner.
If you lower the manifold flowtemp to 40C then the boiler will more than likely trip on a high dT of 33C as its flowrate will have dropped to 3.91LPM.

 

[HRP123]

Your system will still only require 9kw, assuming a manifold dT of 8C then the UFH flowrate will be 16.13LPM at 45C, return 38C with 4.61LPM returning to the boiler giving boiler flow/return flows of 4.61LPM and flow/return temperatures of 65C/38C, 9kw in both but the boilr dT at 27C is getting close to the 30C which normally trips the burner.
If you lower the manifold flowtemp to 40C then the boiler will more than likely trip on a high dT of 33C as its flowrate will have dropped to 3.91LPM.

Thanks for this. Sorry, still self educating about the interaction between flow/return temps and rates ... so are you saying basically that whether the temperature is set to 55 degrees or 40/45 degrees - the boiler will consume 9kw? However the lower flow temperature, leads to a lower return temperature, and if this differential is too large (33) then the boiler will trip and cycle?

 

[John.g]

Yes, regarding the UFH heating requirements and boiler output, both will (have to) be the same. With your system if you want to run at a manifold flowtemp of 40C then you would have to reduce the boiler flowtemp to 59C to give the original 27C boiler dT, normally, you would have a separate UFH manifold recirc pump on each UFH system, then the boiler dT can be controlled to whatever you require irrespective of boiler or manifold temperatures but if your system is working OK then why worry as it certainly wont require more energy and is actually more efficient with those very low return temperatures.
See below schematic of a system with a LLH and UFH manifold recirc pump but with a boiler flowtemp of 65C and UFH flowtemp of 40C and (if) a required boiler dT of 20C, the boiler return will now be 45C.

 

[HRP123]

Yes, regarding the UFH heating requirements and boiler output, both will (have to) be the same. With your system if you want to run at a manifold flowtemp of 40C then you would have to reduce the boiler flowtemp to 59C to give the original 27C boiler dT, normally, you would have a separate UFH manifold recirc pump on each UFH system, then the boiler dT can be controlled to whatever you require irrespective of boiler or manifold temperatures but if your system is working OK then why worry as it certainly wont require more energy and is actually more efficient with those very low return temperatures.
See below schematic of a system with a LLH and UFH manifold recirc pump but with a boiler flowtemp of 65C and UFH flowtemp of 40C and (if) a required boiler dT of 20C, the boiler return will now be 45C.

View attachment 82194

Thanks for this calculation, appreciate it. Sorry for these potentially naive questions, but how do you come up with the LPM rate of 6.45 going into the header and 3.91 going out of the header into the UF circuit? And then the pump LPM rate of 16.13?

 

[John.g]

Start at the UFH requirement of 9.0kw, UFH loops might typically have a dT of 6C to 10C so a dT of 8C might be a reasonable average, now, LPMX60XdT/860 = kw, LPM =9X860/60/8, 16.13LPM. You require a manifold flowtemperature of 40C so the return temperature is, 40-8, 32C. You must mix x LPM at 65C with y LPM at 32C. let LPM32 = flowrate required at 32C, and LPM65 = flowrate required at 65C, LPM65 = 16.13-LPM32, so (16.13x40)=(LPM32X32)+(16.13-LPM32X)X65, solve the equation for LPM32 and you get LPm32=12.22LPM so LPM65=16.13-12.22, 3.91. So you require 3.91LPM at 65C mixed with 12.22LPM at 32C to give you 16.13LPM at 40C and 9kw. You now know that the boiler must suppy that 3.91LPM at 65C and that 3.91LPM at 32C returns to the boiler to give a dT of 65-32, 33C but you might require a dT of 20C so again you use the same method to calculate that you require 2.54LPM at 65C mixed with 3.91LPM at 32C to give you 6.45LPM at 45C to give a dT of 20C and 9kw boiler output.
I do these numbers in a spreadsheet so takes the donkey work out of it, You can see below the LLH part of the first schematic. In the LLH 2.54LPM at 65C is short circuiting in the primary and mixing with 3.91LPM at 32C coming from the secondary to give 6.45LPM at 45C.

 

[HRP123]

Start at the UFH requirement of 9.0kw, UFH loops might typically have a dT of 6C to 10C so a dT of 8C might be a reasonable average, now, LPMX60XdT/860 = kw, LPM =9X860/60/8, 16.13LPM. You require a manifold flowtemperature of 40C so the return temperature is, 40-8, 32C. You must mix x LPM at 65C with y LPM at 32C. let LPM32 = flowrate required at 32C, and LPM65 = flowrate required at 65C, LPM65 = 16.13-LPM32, so (16.13x40)=(LPM32X32)+(16.13-LPM32X)X65, solve the equation for LPM32 and you get LPm32=12.22LPM so LPM65=16.13-12.22, 3.91. So you require 3.91LPM at 65C mixed with 12.22LPM at 32C to give you 16.13LPM at 40C and 9kw. You now know that the boiler must suppy that 3.91LPM at 65C and that 3.91LPM at 32C returns to the boiler to give a dT of 65-32, 33C but you might require a dT of 20C so again you use the same method to calculate that you require 2.54LPM at 65C mixed with 3.91LPM at 32C to give you 6.45LPM at 45C to give a dT of 20C and 9kw boiler output.
I do these numbers in a spreadsheet so takes the donkey work out of it, You can see below the LLH part of the first schematic. In the LLH 2.54LPM at 65C is short circuiting in the primary and mixing with 3.91LPM at 32C coming from the secondary to give 6.45LPM at 45C.

View attachment 82196

Thanks for taking the time to explain all this and share the spreadsheet. Interesting to see how you work out the math above. So how does one work out what the boiler is actually capable of producing in LPM at 65 degrees?

I had a look at the boiler specs but this does not tell me. I'm just trying to work out with lets say multiple circuits on e.g. underfloor and radiators or towel rails or hot water - is the boiler actually capable of producing the total LPM flow from the LLH to these circuits?

Are you a heating engineer yourself? Cheers!

 

[John.g]

If you assume a 30kw boiler and assuming flow/return temps of 65C/45C, dT20C, then this will give it at a flowrate of 21.5LPM (1290LPH) which would be the comfortable flowrate for this output, otherwise the pressure drop across the boiler heat exchanger can get excessive requiring a very high head pump, if your rad demand is say 30kw with flow/return temps of 55C/45C , dT10C, then a flowrate of 43LPM (double the boiler flowrate) is required, the secondary side of a a LLH will achieve this.

No, not a heating engineer. Ex Marine (15 years) and utilities engineer (31 years).

 

[HRP123]

If you assume a 30kw boiler and assuming flow/return temps of 65C/45C, dT20C, then this will give it at a flowrate of 21.5LPM (1290LPH) which would be the comfortable flowrate for this output, otherwise the pressure drop across the boiler heat exchanger can get excessive requiring a very high head pump, if your rad demand is say 30kw with flow/return temps of 55C/45C , dT10C, then a flowrate of 43LPM (double the boiler flowrate) is required, the secondary side of a a LLH will achieve this.

No, not a heating engineer. Ex Marine (15 years) and utilities engineer (31 years).

View attachment 82207

I assume your engineer training has helped with this approach to calculating flow rates etc. Thanks for taking all the time to explain this!

In my case there is an additional pump going from the boiler into the LLH and then from the LLH four heating circuits plumbed off three flow outputs (UF, towel rails, and the HW and CH circuits share one output from the LLH). I don't think the typical radiator demand is anywhere close to 30kw, probably more like 8-9kw typically. HW should be around 8-9kw, UF around 5-10kw (zoned), and TR 6-7kw.

What confuses me though is that in theory the UF demand should be less than 10kw but it typically pulls almost the maximum capacity of the boiler when open (which I have ranged rated down to around 24kw). The heat loss calculations don't suggest this should be the case as the floors and walls are insulated. This is why I wonder if somehow the flow and return (mixed) are distorting something in the LLH ... my very basic noivce understanding is that the UF is pumping out a lot of water from the header (even though only set on pump speed 1) - and/or returning a large flow of cold water - which is reducing the return temp to the boiler for a long period of time, making it think it needs to heat/pump at full capacity (when in reality it could work slower).

 

[John.g]

Most UFH Loops have flow meters thar are adjustable and you can read off the flowrate in LPM, a typical flowrate per loop might be 2LPM with a typical dT of 8C so each loop might emit 2x60x8/860, 1.116kw, say 1.0kw per loop or require a boiler output of 10kw for 10 loops.
Do you have flow readings from yours??, if so and you have flow and return manifold temperature readings then very easy to calculate the total UFH heat output required.

 

[ShaunCorbs]

The ufh will pump out 3 times more flow than the boiler as it’s dt 7 ish

 

[John.g]

What confuses me though is that in theory the UF demand should be less than 10kw but it typically pulls almost the maximum capacity of the boiler when open (which I have ranged rated down to around 24kw). The heat loss calculations don't suggest this should be the case as the floors and walls are insulated. This is why I wonder if somehow the flow and return (mixed) are distorting something in the LLH ... my very basic noivce understanding is that the UF is pumping out a lot of water from the header (even though only set on pump speed 1) - and/or returning a large flow of cold water - which is reducing the return temp to the boiler for a long period of time, making it think it needs to heat/pump at full capacity (when in reality it could work slower).

The LLH only distribute the water in the primary and secondary, they do not lose or gain heat, if the flowrate is higher in one than the other then the dTs will be different to give exactly the same enmergy content.
In the schematic there are three different flowrates and dTs but all have the same energy content of 9.0kw.

The boiler connected with the primary flows 6.45LPm at a dT of 20C, energy content = 6.45x60x20/860 = 9.0kw
The Secondary flows 3.91LPm at a dT of 33C, energy content = 3.91x60x33/860 = 9.0kw
The UFH flows 16.13LPm at a dT of 8C, energy content = 16.13x60x8/860 = 9.0kw

 

[HRP123]

Most UFH Loops have flow meters thar are adjustable and you can read off the flowrate in LPM, a typical flowrate per loop might be 2LPM with a typical dT of 8C so each loop might emit 2x60x8/860, 1.116kw, say 1.0kw per loop or require a boiler output of 10kw for 10 loops.
Do you have flow readings from yours??, if so and you have flow and return manifold temperature readings then very easy to calculate the total UFH heat output required.

I will try and get the flow return temps to the loops as well.

 

[Sparkgap]

How is it controlled? And why does it need to come on in the middle of the night? Just rehashed a local house with 2 rad zones, 2 UFH zones and HWS. Boiler only fires and pumps only come on when timeclocks AND zone/cylinder thermostats are calling.

 

[HRP123]

So I have both timers and thermostatic controls. In this case the UFH came on because the temperature dropped overnight - so it kicked in the heating.

 

[Sparkgap]

I would have thought only the room/s calling for heat shold be getting it, and looking at the photo it would appear that the boiler would come on, quickly heat up the small quantity of water in the LLH and local pipes and shut down. If it is coming on for longer periods and using excessive gas it could be due to unwanted circulation through other circuits, perhaps a motorised valve stuck open or one of the circuits off the manifold not shutting down (or even crossed wires!). I suppose you could wait until everything is timed off and cooled down then override one of the rooms on UFH and see what circuits get warm?

 

[HRP123]

Thanks for the reply! Check all the motorised valves / been watching all the pumps going on and off carefully - so far have not found anything untoward.

My only theory now is with the UFH that the effect of the mixing valve at 50 degrees is somehow distorting the temperature inside the header - such that the boiler thinks it needs to keep firing even though there is sufficient hot water flowing in the UFH circuit.

I'm experimenting range rating my boiler down and using my Nest thermostats to kind of do a hot water priority. In theory it shouldn't make much difference overall but it just means I can control the boiler not going nuts to heat a few circuits because of any distortions in the LLH.

 

[Phill K]

First rule of LLW design. “Thou shall not use multi tapping headers” Sorry but they are a PIN. And if systems arent perfectly matched and flow rates set correctly will leave you wide open to parasitic flow across the ports.

First thing I’d be checking is flow rate of your primary pump, Vs flow rate of your secondaries.

Something is pulling that flow water across the header rather than it circulating around it, so boiler then is ramping up as it’s running on its thermistor readings to control the gas rate. If the heat is going around the system rather than round the header it’s going to take an awfully long time before the temps are coming back to the boiler hot enough to start the modulation process

Guide to low loss headers - CIBSE Journal

Achieving hydraulic separation through well-designed low loss headers is key for efficient heating and cooling systems. David Palmer and Baxi Heating’s Ryan Kirkwood provide detailed design guidance, including a recommended system configuration

www.cibsejournal.com