# where to measure temp on radiator when balancing

Having received my new toy - an IR thermometer from Maplin recently
mentioned here - I set about trying to balance the CH radiators. I
understand that I need to achieve a temp. drop across each radiator of ~
20"F.
The problem I have though is where exactly to take the temp reading
from. The diagrams I've seen show pipe thermometers attached to the
radiator tails but the tails are painted here and therefore somewhat
insulated? I wondered whether taking the temp from the chromed
connections at each end of the radiator might make sense?
Other questions...
How long should I wait after adjusting a lockshield before remeasuring
the temperatures?
The temperature varies depending on the type of surface I point the
laser dot at - do I need to ensure all surface types are the same?
The radiator surface itself reads around 30"F higher than either of the
pipes at each end of the radiator - how can that be?
Sorry for what are probably dumb questions.
You haven't got the type of valve that has both flow and return at one end have you? (Yorkshire Maxtwin)
In message , John writes
Don't think so - Drayton TRV4s at one end of radiator & what is marked as a BS2676-10 lockshield valve at the other.
The message from Si contains these words:
That is the design temperature drop for a particular set of circumstances. What you want is a similar temperature drop across all your radiators.
When faced with the same problem some years ago I opted to take the temperature at the centre of each radiator on the basis that gave a good measure of the average temperature of the radiator.
Until the temperature stops moving but be aware that water temperature in the system is itself a variable and if the house is not up to temperature the return temperature at the boiler may not be high enough to allow the boiler to regulate the flow temperature and if the boiler isn't firing the water temperature is slowly dropping so speed is of the essence in checking all the radiators.
It depends on the emissivity of the surface. According to a table I have oil based paints have a total emissivity in the range 0.92 - 0.96 while for polished copper the figure is 0.052.
In article , Si writes:
Yes (for a conventional boiler), although it more imporant that they are all the same than that they are all any particular value. If one of your radiators forms the bypass (lockshields at both ends), you need to be careful you don't close that one down too much for the boiler.
Pointing at paint is fine. IR themometers don't work well on shiny metal, and copper even when not shiny is particularly bad, as these have low emissivity.
When it stablises. A few minutes probably, but check it a fews times to find out.
Reading from a surface with low emissivity will give an artificially low reading (unless the IR thermometer is recalibrated for that particular surface).
The painted tails are fine. The chromed bits, or bare copper, will give you wildly inaccurate measurements as their emissivity varies considerably from that the IR thermometer is calibrated for.
With an IR measurer - until the readings are stable - usually only 5 minutes or so.
Yes - you can create "equal" surfaces by wrapping a bit of plastic electricians tape around bare metal surfaces to be measured.
Were you measuring from the painted pipes at the end or the chromed surfaces of the valves?
IR "thermometers" measure the amount of Infra Red radiation emitted by the surface rather than their temperature and are designed to work with surfaces with emisivities of around 0.95 (the range is often marked on the device somewhere). Most painted surfaces are close to this but bare metal surfaces are nowhere near and readings from these will not represent their true temperature. You can easily see this effect if you track with the thermometer from the painted tail, over the valve and on to the radiator. Although the true temperature hardly varies the reading on the IR measurer changes a lot as you track over different surfaces.
Also note that at short ranges the laser dot is not where the reading is being taken - it is offset from the sensor and this needs to be taken into account.
In an earlier contribution to this discussion,
As others have said, the emissivity of the surface is important. I have stuck 1" squares of black tape on all my rads - at the bottom, on a flat surface just inboard of the tail - and aim the thermometer at these for consistent results.
Make up a chart, and go round and log the flow and return temps of each rad before changing anything. [Make sure that TRVs and lockshields are fully open before you start, if doing it for the first time - preferably removing TRV heads to be sure they can't operate. Also turn up any room stats to ensure that the system stays on - opening a few windows if necessary].
Having done the initial measurements, start turning down the lockshields on the rads which have the *smallest* temperature difference - doing no more than 2 or 3 at a time. Re-log the whole lot after about 10 minutes - because restricting the flow on some rads will have increased it on others. After a few iterations, you should be able to achieve more or less uniform temperature differentials on all rads.
When they are all the same, consider whether they are the desired value - usually 11 degC (20 degF if you must use antiquated units!) for a non-condensing boiler and, if necessary, adjust the pump speed. [A higher speed will give *less* drop].
Have a look here for useful guidance on how to do it:

In article , Si writes
The simplest spot I have found is the body of the radiator right beside the inlet/outlet ports, they're so close to the pipe itself to be at same temp for useful purposes, usually dead flat so easy to target and the paint should have the right emissivity (if you have any doubts, apply a bit of pvc tape to the rad and measure the tape and the immediately adjacent paint, same reading means no need for tape, different means apply tape as a target).
If in doubt, the inlet is the hotter of the two :-)
No you are looking to equalize the temperature difference between in and out for all the radiators, when operating in a reasonably steady state.
1) Open all TRVS and valves to max. 2) balance the radiator temperature drops 3) Set TRVs as needed.
Be consistent. Do not measure temps off anything shiny. Use the lowest corners of the radiator at each end.
Put plastic sparky tape on chrome radiators.
TRV in operation?
Everyone has to start somewhere. Don't aim for perfection. Pick two or three rads at each round to make adjustments. Don't worry about the actual temperatures or the size of the eventual difference. Just throttle back the two or three with the smallest difference. If you over do it you may find that a radiator that was restricted in round 1 has the biggest "Delta-T" in round two, so you may have to open it back up again.
On most valves 1/2 3/8 1/4 and 1/8 of a turn from closed are the 'active' settings. i.e more than 1/2 turn is about the same as full on.
In message , Ed Sirett writes
That's Ok, all are painted & the same colour I think.
No, it was fully open and water was flowing. I think that was down to the differing surfaces' emissivity and my unknowingly sloppy description - the lower readings were from the shiny ?chrome fittings. :-/
Thanks to all for your responses. I wanted to get on with it tomorrow but my beloved one fears I may break the CH by so doing and that's extremely ungood at the weekend. So next Wednesday it has to be. Hmph!
In message , Roger writes
As in, instead of aiming for an equal temperature drop across each radiator you went for getting an equal temperature at the centre of every radiator? Are there drawbacks to that approach?
So I need the windows open to ensure all the TRVs are fully open but not so much that the boiler can't cope. Couldn't I just take the tops off all the TRVs?
Hence my silly readings - thank-you.
In message , Peter Parry writes
there's a chart on the side of it - I'll have to find a magnifying glass to read it though.
In message , Roger Mills writes
Sounds like a lot of exercise - do me some good. :)
I was 13 when the Uk decimalised its currency. My education was initially in imperial units and then metric. I tend to use whichever unit suits the purpose. Here it made no difference and 20 was easier to remember than 11. :)
Thanks.
In article , Si writes:
I was the last year in my primary school which learned to do division in £sd...
In message , Andrew Gabriel writes
Wonderful process. :) Today's kids would run for their calculators and then be equally stumped.
I found an old price ticket recently, for '3 bored demi-john corks', priced in the particular typeface Boots then used as |/-
My son & his friends didn't have a clue. :)
In article , Si writes
Not my favoured approach certainly.
Max setting on a TRV4 corresponds to 30degC which is very unlikely to be reached so I would just leave them on but set to max. The balance will be better if the room temperature matches the design control temperature (say 18 or 21degC) so opening the windows to maintain that temperature would be a good idea as it can get a bit hot if you take your time and (by definition) have your radiator controls overridden.
In an earlier contribution to this discussion,
You probably need to do both! You *do* need to keep the TRVs open - and the best way to do that is to remove the heads.
But you *also* need to ensure that water of more or less constant temperature (controlled by the boiler thermostat) is circulating for the whole time you're balancing - otherwise you'll get some very funny results!
With the system running constantly, the house could conceiveably get up to a temperature where the room stat turns it off - which you don't want to happen! Even with the stat turned right up, it will still cut out at about 30 degC - so you may need to open the windows to prevent this happening.
On 4 Jan,
I chucked my first one away (IIRC a Sinclair scientific) when I realised I was getting it out to divide by 10.
That usually refers to the sensor detection area with range. Think of it as a torch with a not very good reflector - as you move away the area the sensor covers increases. This means that if you are trying to get the temperature of a hot pipe against a wall you need to be quite close so the sensor area is no bigger than the pipe diameter otherwise you get a reading comprising part wall and part hot pipe which will show as hotter than the wall and cooler than the pipe.

• how to remove emulsion from the ceiling, its flaking, im told because under it...
• next in

## UK Do-It-Yourself

• Hello. I have MFI contracted installers in at the moment doing installing an...
• previous in

## UK Do-It-Yourself

• I wanted to measure to consumption of a 12V pump. It was obviously going to be...
• last updated in

## UK Do-It-Yourself

• I hear a lot of those "working from home" want to make it their normal practice...
• last posted in

## UK Do-It-Yourself

• Soooooo since no one is mentioning building something I'll mention the POS I...
• site's last updated in

## Woodworking

• THE SNITCHING BEGINS: NOSEY NEIGHBOUR CALLS POLICE ON WAKE IN PUB GARDEN ?There...