Radiator balancing problems
Hi there
I have been trying to balance my radiators without much luck, so I am hoping someone can help tell me what I am doing wrong :s
I have the Evohome system with 10 radiators - six with HR92s and 4 with manual controllers.
I've read pretty much every article and forum post there is on the subject over the last week, but I am still struggling.
I opened up all the lockshields and controllers, determined the order they heat up, then went round each radiator in order (except I only got as far as the second), closed off the lockshield, then opened it slowly to try and achieve close to the 12 degree drop.
I bought two Brannan pipe thermometers for the job.
Firstly, it took me so long to try and balance the first two radiators that the rest of the house was already 24+ degrees so I gave up.
First radiator was 44 in / 35 out with the LSV @ 1/2 turn, and after a lot of trial and error, I got it to 44/32 with LSV @ 1/4 turn.
Second radiator was 42/42 with the LSV @ 1/4 turn, 42/37 with LSV @ 1/8, and 43/34 with LSV @ 1/16.
So I stopped and tried the next day - starting at the second radiator. However after spending an hour trying to adjust this one I realised that the first radiator I adjusted previously was still cold! I've opened that one back up to 1/2 turn now..
Not sure where I am going wrong. I am wondering if I got the order of the radiators wrong. I thought the hallway was first (as it is closest to the boiler), then cloakroom, then lounge, then upstairs.
However I think maybe the bathroom is first as it is directly above the boiler. It's hard to tell because when I turn the heating on, they start to feel warm at about the same time.
I was thinking of investing in an infrared thermometer if this will make it easier, but that still doesn't get around turning the house into an oven every time I try...
Many thanks in advance
Mike
Just a thought but perhaps you don't need to balance them? I may be that if the 4 with manual controllers had HR92s and all lock shields were left fully open it would be fine. I have never had to do any balancing of mind, perhaps i ought but have left them to be controlled by the HR92s and have no issues.
Balancing radiators properly is very time consuming and does involve a lot of adjustment and readjustment. Last time I did our system it took a good few hours. (I am a perfectionist mind you!) Part of the reason for that is that you have to let the radiators cool down again after an initial adjustment to test how well the system is balanced next time it warms up, to make further adjustments to refine the balance.
It's probably quite a bit quicker to use a non contact IR thermometer, I have one similar to this:
http://www.ebay.co.uk/itm/Digital-IR...3D231611680353
You can point it at any part of a radiator and get an accurate reading in a couple of seconds whereas contact thermometers take time to settle on the final reading. I actually bought mine for use on cars (measuring exhaust temperature etc) but found it handy for measuring central heating radiators and frying pans!
Where you might be tripping up is obsessing too much about getting an exact 11/12 degree temperature drop across the radiator. This figure is quoted nearly everywhere you read online but the truth is it's only a rough rule of thumb and it's not applicable to all radiators.
Heat loss across a radiator doesn't just depend on the flow rate it also depends on how well the radiator does its primary job - radiating heat. If your radiators are all the same type and are all convector's its a reasonable rule of thumb, but if you have a mixture of old and new radiators like we do, especially where some radiators have convector fins and some are old fashioned types that don't, trying to balance them all to 11 degrees will only lead to madness, and may actually be impossible...
For example when I tried to balance ours all to 11 degrees I found that two of our old, small, non-convector radiators could not be closed off enough to get an 11 degree drop without virtually closing the lockshield valve - at which point the overall temperature of the radiator was some 20 degrees or more below the flow temperature.
EG the flow temperature (and the hot side of most other radiators) might be 70 degrees, but the hot side of this radiator was only 50 degrees and the cold side 40 degrees. Opening the valve up enough to bring the hot side of the radiator up to 65-70 degrees would also bring the cold side up to within 3-4 degrees of the hot side.
The reason is quite simple in hindsight - a single panel non-convector that's 800x740mm can't radiate enough heat to get an 11 degree temperature drop at any reasonable flow rate. In my opinion that 11 degree rule of thumb is only valid if all your radiators are modern convector's and don't differ too much in size.
We later replaced this old radiator in the bathroom with a modern 800x700mm dual panel single convector and the difference is astounding - now I actually have to have the lockshield valve fully open just to bring the temperature differential down to 11 degrees, and it only just makes it. The bathroom can get a lot hotter now too.
I'm no expert but in my understanding the point of balancing radiators is not to strictly adhere to an 11 degree rule of thumb, but adjust the pressure balance between the radiators such that when the TRV's are all fully open and the system is starting from cold, all radiators warm up at roughly the same rate. What you don't want is to have a couple of nearby radiators that get hot really quickly while far away radiators stay warm or even cold until the TRV's on the nearer radiators close.
By partially closing the lockshield valves on those near radiators you increase the flow resistance through those nearer radiators so that it more closely matches that of the further away ones. In theory with the right adjustment you could get all the radiators warming up at about the same rate, eg same average surface temperature during the initial warmup phase, and that should be the goal of the exercise. If the radiators differ in type greatly then they won't all have the same 11 degree differential.
I suspect the oft quoted 11 degrees is more about wanting to reduce the return flow temperature so that a condensing boiler can run in condensing mode more often, thus run more efficiently.
However if the differential is too small on all radiators then rather than trying to close all the lockshield valves down too far you probably should be reducing the speed of the pump or possibly adjusting the automatic bypass valve. The lockshields are for adjusting imbalance between different radiators, not for trying to restrict the flow on all radiators if the pump speed is too high - that will only lead to the automatic bypass valve opening (if you have one) which will then result in a high return temperature to the boiler and a poor efficiency on a condensing boiler.
So, I wouldn't worry too much about the exact temperature drop - I would focus on getting the radiators warming up at the same rate during the warm up phase, then when they're all hot measure the temperature differential across each radiator to see if it's "reasonable" - you don't want it to be any more than 12 degrees on any radiator but if you have a couple of small or non-convector's where the drop is only 4-5 degrees I would not worry about that at all.
If the differential is too low on all radiators and correcting it would require the valves to be nearly closed I would try dropping the pump speed a bit instead and then go through the balancing process again.
Hope that helps.
Last edited by DBMandrake; 24th February 2016 at 09:33 PM.
Second everything from DBMandrake. I installed my own heating system many over 25 years ago but added some additional rads a few years ago. Rather than the original set up like you did that took hours, "I learnt" the following
1) Not all rads can achieve the ideal temp drop - I have some towel rads that only achieve a few degrees despite spending hours trying to get the full temp drop
2) Use an IR detector and indeed I have exactly the same one in the link - they are soooooo much quicker than attaching sensors whether mechanical or electronic to the pipe and waiting for the temp change
3) The temp differential only applies when the boiler is heating. If cycling (ie the boiler is not actually firing as it is attempting to disperse the heat)you have to wait for the next reheat otherwise the difference between inlet and outlet narrows during the cycle
4) Set the lockshield valve on the furthest rad fully open and reduce them as you get closer to the boiler and I tend to have the rad closest to the boiler just open enough to just allow water through and vary everything in between by approximately the number of rads
5) Start with the pump speed on full and if with this crude setup all rads heat up try reducing the speed (unless it is an auto speed pump) until you get the slowest speed that the rads all heat up with - you will need this for the reasons stated above as you don't want to overload the pump as rads shut down and also it can be very noisy
6) Now balance the rads using the IR detector (start at the nearest to the pump) briefly pointing at the inlet pipe and outlet and tweak the lock shied valve until you get a temp drop but if you cant get it down to the 11 degree then don't sweat it. There needs to be some temp drop but the main thing is that all rads heat up evenly and roughly in step. Go onto the next one and repeat.
7) if you find you are closing lockshield valves you may need to reduce the pump speed. If you find you are opening lockshield valves you may need to up the pump speed.
Oh and it all works easier on a cold day and a cold house. A hot day and a hot house wont allow the rad to disperse the heat it is designed to so again that is why the 11degrees is a design theoretical and difficult to achieve in reality.
One suggestion I'd make in the use of an IR gun is don't point the IR gun at the actual copper pipe, for two reasons:Originally Posted by SteveP
1) An IR gun has a finite spot size which it averages received IR over. This means that to measure something as narrow as a 15 mm radiator pipe you would need to hold the opening of the gun so close that it's nearly touching otherwise the reading will be lower because some of the viewing aperture of the gun is looking at IR from colder objects behind the pipe - for an accurate reading all the IR received by the sensor has to be from the object under test. On 8mm microbore you can't hold the gun close enough to the small pipe to get an accurate reading.
2) The emissivity of polished copper (or any shiny polished metal for that matter) is very low, this means that the "temperature" reading given by the IR gun (if you leave it on the default 0.95 emissivity) will be much lower than the actual temperature - it could read as low as 40 degrees when the pipe is really 70 degrees. While most IR guns lets you set the emissivity calibration to suite the material, scuffed copper has much higher emissivity than polished copper so it's nearly impossible to estimate what it's emissivity might be based on the degree of smoothness/scuffing.
On the other hand paint typically found on radiator panels has a nice predictable emissivity of about 0.95 and is very easy to get accurate readings from. The panel is also plenty large enough to fill the entire field of view of the IR sensor. So I take the readings from the bottom corners of the panel - holding the gun about 2-3 inches from the panel aimed at a point an inch or two from the bottom and side edges. Move the gun around slightly to find the highest temperature at the corner and use that. (This assumes your pipes connect to the bottom corners - if not measure on the corners where the pipes connect of course)
If you have 15mm pipes that are painted with matt paint you can probably get a decent reading off the pipe if you hold it right up but I would still suggest measuring close to the corner of the panel will give more accurate repeatable results.
Last edited by DBMandrake; 24th February 2016 at 11:52 PM.
I also struggled with this. Here's how I eventually figured out how to do mine. Rather than getting it all balanced in one session, I just gradually worked towards it over a period of a couple of weeks.
Open all radiators a couple of turns, enough to ensure that each radiator gets nice and piping hot -- set everything to 24 degrees and within 30 minutes all the radiators should be almost too hot to touch. The first thing you want to do is confirm that all the radiators are performing fully, no sludge (cold at the bottom) or air (cold at the top).
The key for checking if you are balanced or not, is that the output valve should be perceptibly LESS-HOT than the input valve. Just using your hand, grab the input valve, and it should be scalding hot (assuming your boiler is set to 72-ish Celcius). Now grab the output valve, and it should still be hot, but definitely not as hot as the input valve (by definitely, I mean that there should be no doubt in your mind that it is less hot. If you doubt, it is still too hot).
So... If the output valve is luke-warm, you need to open the valve a little more (tiny increments, no more than a quarter-turn in any given "fiddling session"). If the output valve feels just as hot as the input valve, then you need to close the valve (no more than a quarter-turn). Repeat this every time you remember over the normal course of the days and weeks ahead. The key for these gradual tweaks is this: both the radiator and the input valve need to be piping hot (this will only happen when the heating comes on after a period of being off, e.g. early in the morning, or when you get home from work). If both those things are true, then feel the output valve, and adjust as above. Eventually you will reach a point where the output valves are all less hot than the input valve, and at that point I would consider the system balanced.
In summary: if both the radiator and the input valve are piping hot, then the output valve should be hot, but not piping hot. If the output valve is lukewarm then you are heating this radiator too slowly or not enough; if the output valve is piping hot then you are heating this radiator too fast.
My panels (even in the bottom corner next to the output valve) can be piping hot, and the output valve can still be lukewarm. So measuring the panel is not a good measure in my case.
Just a note to say I appreciate all this info re balancing. Not something I considered before.
That can only be the case if the boiler is no longer firing and/or the pump has stopped circulating, because the pipes and valves will cool much faster than the bulk of the radiator panel.
As SteveP pointed out in his post you can only measure the differential temperature across the radiator when the boiler is running continuously. The lower reading you get on the pipe is invalid because either no water is flowing or the water that is flowing is cool because the boiler is not firing.
This is why the ideal time to measure the differential temperatures is during the initial warmup phase where the flow temperature is not yet quite up to the set temperature thus the pump and boiler are running continuously.
I was surprised this was the case, but definitely happens while water is still flowing. Maybe the output comes from the top of the radiator and not the bottom?
Reply With Quote
