Oversizing Rads for lower circulation temperature

[flappy8]

Hi all,

Is there a rule of thumb for increasing the radiator size with lower circulation temps? My house is going to have wet UFH downstairs and rads upstairs and I was thinking about getting the circulation temperature at around 65c. What percentage increase of rad size should I allow as compared to 80c flow?

 

[Worcester]

No, it's better than that! There are proper, calculated formalae

What flow temperature are you looking for

Most Rads are sized / specified for a DeltaT of 50°, i.e. average flow 70° = 75° flow 65° Return

So based on standard rad sized for DeltaT of 50

Flow Radiator increase Factor
35° 6.8
40° 4.3
45° 3.1
50° 2.4
55° 1.9
60° 1.6

So for 60° flow it should be 1.6 times bigger.

Hence when we design Heat Pump systems we are looking for a balance between flow temperature ( = Return on Investment as CoP goes down with increase in flow temp) and Size of Radiator, hence normal design flow is 45° for a retrofit heat pump as rads that are 3.1 times bigger (using triple panels / K3's etc) are easy enough to accomodate.

 

[flappy8]

No, it's better than that! There are proper, calculated formalae

What flow temperature are you looking for

Proper formulae are good!
Thanks Worcester – I was hoping to get the flow temperaturedown to 60 to make it more efficient but give me hot water. So now I know what I need to look at rad wisein the smaller upstairs rooms.
Thanks

 

[lame plumber]

most houses have rads way oversized anyhow as plumbers are lazy/wary by nature and wack them in to fit the window or wall. why not measure up what you have already and then do the clcs, bet you dont need to touch them.

 

[Ermintrude]

Massively oversize. Out each room or cluster on a zone valve or actuator and have a stat in each room it's what we are doing a lot more. John guest manifolds can be used like this. I prefer to build my own with Honeywell two ports as infinitely more repairable

 

[oz-plumber]

The other thing to take into consideration is that if you have heating on the ground floor, you ( roughly ) only need 75% of the calculated heatload for rooms on the first floor.

Chances are what you have got already will be sufficient.

As per previous 2 posts

 

[chris watkins]

The other thing to take into consideration is that if you have heating on the ground floor, you ( roughly ) only need 75% of the calculated heatload for rooms on the first floor.

Chances are what you have got already will be sufficient.

As per previous 2 posts

Sorry oz but where is this magical 25% going to come from then on the upper floor? If the air temp is the same between the rooms on the upper & lower floors there will be no heat flow between them.

Is not the output of a emitter also not based on the room temperature not just the Mean Water Temperature so ΔT 50 = (75+65÷2=) 60 MWT - 20°C Room temp = 50
The areas of UFH may be designed to 20 but most rooms heated by Rads are 21, 22 for Baths & Showers, 23 for elderly (halls & bed could be designed to be as low as 18°C)

That is not to say that your "increase factor" is not very useful though Mr Worcester but it is good to understand the whole picture.

 

[Worcester]

Is not the output of a emitter also not based on the room temperature not just the Mean Water Temperature so ΔT 50 = (75+65÷2=) 60 MWT (70 Corected)- 20°C Room temp = 50

True room temps are +- around 20 and if you want to get more accurate you can, however heat loss calcs even to BS EN 12831 don't take allowance for solar gains etc)
So I'm not worried about that small difference

As for the increase factor - it isn't mine It is provided by the industry who all done their extensive thermodynamic calculations and modelling - it's all included in chapter 10 of the CIBSE Domestic Heating design guide and the supplement.

If you want to be more accurate you should also include a factor for how you've connected the pipe, (TBSE, TBOE, BOE) how it's fixed - is there a shelf over it? and what it has been painted with you could then also add in allowances in the room for any exposed pipework (usually ignored in domestic environment, possibly a major heat outpt / loss in a commercial environment)

The factors above are simply provided by the industry and based on tables 10.1, 10.2, 10.3 and 10.4

Hope that helps.

 

[chris watkins]

True room temps are +- around 20 and if you want to get more accurate you can, however heat loss calcs even to BS EN 12831 don't take allowance for solar gains etc)
So I'm not worried about that small difference

As for the increase factor - it isn't mine It is provided by the industry who all done their extensive thermodynamic calculations and modelling - it's all included in chapter 10 of the CIBSE Domestic Heating design guide and the supplement.

If you want to be more accurate you should also include a factor for how you've connected the pipe, (TBSE, TBOE, BOE) how it's fixed - is there a shelf over it? and what it has been painted with you could then also add in allowances in the room for any exposed pipework (usually ignored in domestic environment, possibly a major heat outpt / loss in a commercial environment)

The factors above are simply provided by the industry and based on tables 10.1, 10.2, 10.3 and 10.4
View attachment 18862
Hope that helps.

You just made me read through my new copy of DHDG 2013, I had not noticed that new table (10.1) for the correction factor you posted, it is probable me but it took a while to work out how to use it unlike the older copies where you just found, say, 39 & it showed 0.724. In this one you have to look down the left hand column find the 30's & then move across to the 9 to read off 0.724.

If you know the heat loss you require, you divide this by the factor given & use this to select a BS EN 442 rad, it will then provide the heat output you require given the different temperature of water it will be supplied with.

 

[Nostrum]

Is it just me or do we have some extreme willy waving going on here ?!

whats changed in the latest CIBSE, does it warrant buying it?

 

[chris watkins]

Is it just me or do we have some extreme willy waving going on here ?!

whats changed in the latest CIBSE, does it warrant buying it?

Absolutely no idea what you folk get up to down in deepest darkest Devon but up here we keep our appendices firmly in our pants & wave only our :waving:'s !!!!!

RE the DHDG, IMHO you can make do with the 2007 copy but it is worth upgrading to the latest to stay ahead of the game with the changes to Part L & F etc.

 

[oz-plumber]

Sorry oz but where is this magical 25% going to come from then on the upper floor? If the air temp is the same between the rooms on the upper & lower floors there will be no heat flow between them.

Basically Chris,
The ground floor temp as you stated is @ 20C.
The temperature required to heat the first floor will not have an ambient temperature base of ( in my case 3.5C) and in your case @ -1 C.
Unless the occupiers of the dwelling have the UFH system turned off completely, and live on the first floor then the 'roughly adjusted factor' of 75% will be sufficient to heat the first floor to a desirable temperature.

Of course the figure of 75% is an average, depending on type of construction and insulation factors, windows and so forth.

From my experience, and the experience of my customers, the base setting for the UFH system would be 18C and when the home is occupied @21C
.
So technically speaking, due to the response times of UFH the lowest temperature set on the thermostat is 18C, this would enable a much lower heatload on the first floor due to not having to base the heatload calculations on the outside ambient temperatures.

This factor may change if a room is built above a garage or over an area that is not heated by the UFH system.

Oz-Plumber

 

[chris watkins]

I am not sure I understand much of the above in terms of the thread OZ, the basics of heat transfer is there will be a flow of heat energy from hot to cold & each surface area (internal & external walls, floors & ceilings) needs to be considered.

When sizing heat emitters a calculation must be done to work out how much heat is going to be losses or gained by the space that the emitter will serve.
If the room on the first floor is at a higher temperature than the one directly below it then it will loose heat in that direction & the rad, lets say, will have to be larger to allow for this whereas the room below would have a smaller UFH system because it is gaining the heat from above.

It works like this wherever there is a difference it temperature whether we are talking inside or outside of a building.

P.S. We normally use -3° C but it does depend on the part of the country you are in a height above sea level.

 

[lame plumber]

I always understood that hot air rises, so how does a first floor room heat a lower floor room please

 

[chris watkins]

I always understood that hot air rises, so how does a first floor room heat a lower floor room please

Did you get that old nugget about hot air rising from school lame ?? Well if so, they lied to you! Teachers tend to do that when they can't be asked to explain things properly.

Hot air by itself does not rise, it is pushed out of the way by the colder air around it. It is the weight of the cooler air being more dense that pushes the lighter warmer air up.

Now to your question, we are talking about conducted heat transfer here, not convection, taking that the floor / ceiling between the two rooms is reassemble air tight there would be no air movement between them.
You seem to be saying that there can't be heat flow downwards but lets think about it, if that were the case a floor at ground level should not feel cold to the touch but during the winter it does !! If there is any difference in temperature between two surfaces then heat will start to conduct from the warmer to the cooler so it will feel cold to the touch, remember it can't move the other way.

The greater the difference in the temp, the greater the heat energy flows.

 

[newcastle phill]

Is this april the 1st

 

[josh]

As this is a good science lesson from Chris I'v got a couple of questions,

Some indirect cylinders now have a coil rating of 20kw but you obviously don't add that on top of a boiler sizing if not got hot water priority nor will you get 20kw through a 22mm pipe so how have they worked that out? I know it's to do with surface area but someone that knows a bit about it would be good to explain.


Also why when you run a hot tap and have to draw of some cold first does the noise change when the water starts to run hot?

 

[chris watkins]

As this is a good science lesson from Chris I'v got a couple of questions,

Some indirect cylinders now have a coil rating of 20kw but you obviously don't add that on top of a boiler sizing if not got hot water priority nor will you get 20kw through a 22mm pipe so how have they worked that out? I know it's to do with surface area but someone that knows a bit about it would be good to explain.
Also why when you run a hot tap and have to draw of some cold first does the noise change when the water starts to run hot?

Hi Josh & welcome to "Science on Sunday" LOL
You are absolutely correct, most cylinders now have heat exchangers around the 20kW mark, this is to try to meet the requirement of Part L to reheat within 20minutes & also deliver a 20deg C drop when cylinder is heating.

You are also correct that this load should not be added to the heating load when sizing the boiler. If your system does not have HWS priority then clearly the boiler is not going to be able to heat both simultaneously with satisfactory results so it makes sense to time the HW at least half an hour before the heating is required (at least is thing in the morning).
You are incorrect though with the carrying capacity of 22mm tube, it can carry around 23kW @ a 20deg C drop, I guess thats why I have not seen a cylinder with 22mm con's with a coil larger than 22kW. (it is all about the surface area of the coil)

Regarding the hot water tap noise, I haven't got a real answer, if I was to hazard a guess, I would say it is due to the expansion of the metal or more likely the washer as the hot water arrives changing the rate of flow slightly??
Anyone else got any ideas ?

 

[doitmyself]

most cylinders now have heat exchangers around the 20kW mark, this is to try to meet the requirement of Part L to reheat within 20minutes & also deliver a 20deg C drop when cylinder is heating.

... the carrying capacity of 22mm tube, it can carry around 23kW @ a 20deg C drop, I guess thats why I have not seen a cylinder with 22mm con's with a coil larger than 22kW. (it is all about the surface area of the coil)

Just had a thought!

1. Heat always transfers from the higher temperature to the lower.
2. While the water in the cylinder is lower than that flowing though the coil, heat will be transferred from the water in the coil to the water in the cylinder.

Question: Does that mean that the temperature of the water leaving the coil can never be lower than the temperature of the water in the cylinder? (If it could, heat would get transferred from cylinder to coil.)

If true, to achieve a cylinder temperature of 60C with a 20C drop will require a flow temp of 80C.

 

[T & H Plumbing]

its like being back at College wi all this jibba jabba :33:

good posts thou

 

[chris watkins]

Just had a thought!

1. Heat always transfers from the higher temperature to the lower.
2. While the water in the cylinder is lower than that flowing though the coil, heat will be transferred from the water in the coil to the water in the cylinder.

Question: Does that mean that the temperature of the water leaving the coil can never be lower than the temperature of the water in the cylinder? (If it could, heat would get transferred from cylinder to coil.)

If true, to achieve a cylinder temperature of 60C with a 20C drop will require a flow temp of 80C.

Your question is based on there being a uniform water temperature across the secondary water inside the cylinder but as the water around the outside of the coil is heated it floats upwards until the cylinder reaches the temperature set on the thermostat at that point.

You are correct, if the cylinder secondary water is hotter than the primary water heat will be carried away.

Modern cylinders with there large coil heat exchangers are designed so that when they start to be heated, the return water temperature is as low as possible this helps the condensing boiler stay in condensing mode & nicks a bit more energy out of the flue gases. The large drop also means that the boiler can work out how much gas it should be burning (modulation) without switch on & off (cycling) which wastes gas.

This large drop, say your 20°C, cannot be sustained as the secondary water starts to reach that of the primary water, so boiler goes out of condensing mode but the job has been done.
The primary water temperature must be at least 5 -10°C above that required in the secondary otherwise it will take for ever to finally reach the desired temp. (Primary flow of 70°C is now recommended for indirect systems).

 

[Ermintrude]

ACV sle 300 can lift 70kw. Best not connect that up to a 22 primary! Coils are in the past. Jackets are way forward as they can be used as a buffer

 

[doitmyself]

Your question is based on there being a uniform water temperature across the secondary water inside the cylinder but as the water around the outside of the coil is heated it floats upwards until the cylinder reaches the temperature set on the thermostat at that point.

Yes , I was simplifying the problem by ignoring the convecton currents within the cylinder.

You are correct, if the cylinder secondary water is hotter than the primary water heat will be carried away.

But at least I got that bit right.

This large drop, say your 20°C, cannot be sustained as the secondary water starts to reach that of the primary water, so boiler goes out of condensing mode but the job has been done.

Which will be at approx 55C.

The primary water temperature must be at least 5 -10°C above that required in the secondary otherwise it will take for ever to finally reach the desired temp. (Primary flow of 70°C is now recommended for indirect systems).

That sounds a sensible compromise.

Presumably you balance the primary for a 20C drop when the water in the cylinder is cold.

 

[chris watkins]

Presumably you balance the primary for a 20C drop when the water in the cylinder is cold.

I don't believe in balancing primary,doit, I think the days of trying to do heating & hot water simultaneously are gone. So time HW to come on first & then give it everything the boiler has for 40mins then over to heating (first thing in morning that is, rest of the time its just top-ups if required).

 

[doitmyself]

I don't believe in balancing primary,doit, I think the days of trying to do heating & hot water simultaneously are gone. So time HW to come on first & then give it everything the boiler has for 40mins then over to heating (first thing in morning that is, rest of the time its just top-ups if required).

But if you have only the cylinder in circuit the head seen by the pump will be very low so, assuming a fixed speed pump, the flow rate will be high and the temperature drop consequently low, maybe only 5C or so. Wouldn't it be better to restrict the flow through the cylinder so you get a drop of 20C to start with, i.e when the DHW was cold?

 

[Ermintrude]

I don't believe in balancing primary,doit, I think the days of trying to do heating & hot water simultaneously are gone. So time HW to come on first & then give it everything the boiler has for 40mins then over to heating (first thing in morning that is, rest of the time its just top-ups if required).

U can wire s plan with cyl stat closing the ch two port

 

[Ermintrude]

But if you have only the cylinder in circuit the head seen by the pump will be very low so, assuming a fixed speed pump, the flow rate will be high and the temperature drop consequently low, maybe only 5C or so. Wouldn't it be better to restrict the flow through the cylinder so you get a drop of 20C to start with, i.e when the DHW was cold?

When the DHW is cold the rate of heat transfer will be it's best so deliver as fast and as cool as possible. What's really needed is a proportional temp modulation controller

 

[oz-plumber]

I am not sure I understand much of the above in terms of the thread OZ, the basics of heat transfer is there will be a flow of heat energy from hot to cold & each surface area (internal & external walls, floors & ceilings) needs to be considered.

When sizing heat emitters a calculation must be done to work out how much heat is going to be losses or gained by the space that the emitter will serve.
If the room on the first floor is at a higher temperature than the one directly below it then it will loose heat in that direction & the rad, lets say, will have to be larger to allow for this whereas the room below would have a smaller UFH system because it is gaining the heat from above.

It works like this wherever there is a difference it temperature whether we are talking inside or outside of a building.

P.S. We normally use -3° C but it does depend on the part of the country you are in a height above sea level.

Hi Chris,

Are you saying that the temperature at floor level, on the first floor, will be higher than the temperature at the ceiling level of the ground floor - assuming that the UFH system would be operating, or producing heat, for the majority of the time when heating is required.

There would be heat transfer from the room below to the room above.

Are you also saying that when you design/size a heating system for a dwelling that has ground floor and first floor, do you size the first floor using the same parameters as you use for the ground floor?

Oz-Plumber

 

[chris watkins]

But if you have only the cylinder in circuit the head seen by the pump will be very low so, assuming a fixed speed pump, the flow rate will be high and the temperature drop consequently low, maybe only 5C or so. Wouldn't it be better to restrict the flow through the cylinder so you get a drop of 20C to start with, i.e when the DHW was cold?

I take your point doit but with most domestic systems the coil will be rated higher than the boiler max output feeding it so it will be able to take all the flow rate the system has.
No need for balancing valve.

 

[kay-jay]

Hi Chris,

Are you saying that the temperature at floor level, on the first floor, will be higher than the temperature at the ceiling level of the ground floor - assuming that the UFH system would be operating, or producing heat, for the majority of the time when heating is required.

There would be heat transfer from the room below to the room above.

Are you also saying that when you design/size a heating system for a dwelling that has ground floor and first floor, do you size the first floor using the same parameters as you use for the ground floor?

Oz-Plumber

Yes you use the same parameters of heatloss yo heat upstairs. You have to take into consideration heat transfer from high to low as well though. This forms part of heatloss and has its own u values depending on the temp difference between the two.
In the uk upstairs rooms are likely to be bedrooms which are designed around 3 degrees cooler than living spaces downstairs. So when calculating heatloss gor upstairs take into account heat transfer (gain in this example) from the warmer lower floor.

 

[chris watkins]

Hi Oz, as per key-jay post above, although I might disagree with the design temps he suggests (see post 7). The point is really that you need to decide the steady room temp you require what ever it is, there is no point in saying it is hot or cooler at different levels cos it is a dynamic system.

You take this design room temp & then look at every surface of that room, walls, ceilings, floors etc & consider what is on the other side if it is hotter, heat will be gained to the room, if it is cooler heat will be lost.
You would do the same for every room in the house & would arrive at a loss for every room & so would be able to accurately size each heat emitter to provide the heat required to each (plus any increase % for intermittent heating, warming up after it has been turned off)

Add them all up & you can size the boiler required.

 

[Ermintrude]

Or just look at gas supply and guess what it will feed? 15mm off old icos....... Well hard to tell icos never burned gas........ Probably 40kw open vent love wiv powershowers pumps like?

Building a spreadsheet is a good way to do this but an excellent way is to use Matlab and build functions inputs then model it and it's pretty interesting - bit oTT for a hoose! Spreadsheet is dynamic enough.

Don't forget to factor in air changes tho !

 

[chris watkins]

Don't forget to factor in air changes tho !

Oooop's thanks Ermi, in all the excitement I forgot the ventilation losses.
Air changes per hour x room volume x 0.33 x temp diff = add this to fabric loss, then job done.

 

[mutley racers]

Wow, that's a bit of an education

 

[Worcester]

What an excellent thread

This should become a sticky methinks

Question: Does that mean that the temperature of the water leaving the coil can never be lower than the temperature of the water in the cylinder? (If it could, heat would get transferred from cylinder to coil.)

Yep, that's why we put non-return valves in solar thermal systems, else they could themosyphon the heat away!