Total confusion over balancing radiators

[Percy]

I'll try to keep this as short as possible. My heating system is 28 years old and is 8mm microbore with standard steel panel radiators and an indirect hot water cylinder.

I spent a couple of weeks this summer totally overhauling the system by removing all the radiators and flushing them, flushing each individual pipe in the system with (cough) mains pressure water in both directions at the pump attachments and fitting 15mm radiator valves with reducers to 8mm.

Quite a lot of gunk came out of the rads and the pipe work and now that the system is refilled and bled the rads get hot much quicker than before.

It's the balancing that I'm having problems with. I've read all the info I can find and frankly none of it seems to have the desired effect on my system and there are other problems.

My pump has 3 speed settings and the middle setting seems to work best.

I started by turning all the main valves and TRVs fully open and the lock shield valves to a quarter turn from off and immediately noticed quite noticeable sounds of rushing water in a few of the radiators. If I open the lock shields the noise goes away but I now have no control over the flow through the radiators.

I am trying to use one of these

to measure the temperatures but I can't get a reading from the pipes as they are only 8mm and trying it on the bottom corners of the rads gives wildly varying results. I have heard mention of putting a patch of black tape on the rads but, is this really necessary?

I can't get near to the required drop across the rads and soon get back to the point where the rushing sound returns before any difference is noticeable.

What am I doing wrong and what should I be doing?

[Dave Liquorice]

Is the 8mm pipe work connected to a manifold(s) somewhere? I suspect it is rather than being a loop with the rads connected in parallel across it. I'm not sure how one would go about balancing a manifold based system, does the FAQ cover it?

Noise indicative that the water is being pushed too fast through a small gap, turn the pump down. Also with the water circulating fast it doesn't get a chance to lose heat to the room, hence not able to get the 10 degree or so drop across a given rad.

[fred]

In article , Percy writes

Is this the method you are following:

My pump has 3 speed settings and the middle setting seems to work best. >

That's a good start, yes.

Classic noob mistake, what you have done is throttle the system and that is not what you do at all. You start with all lockshields fully open and only start to close ones that show low temperature drop. At the end of the process (after many cycles) you will expect to see lockshields on large rads on long pipe runs far from the boiler still fully open and small rads close to the boiler perhaps on the quarter or half turn open that you have attempted. Use a measurement table as described and measure all rads within 5 or 10 mins, then work out how much each is to be adjusted on that loop and make all the changes at once before waiting

20mins for the system to stabilise again. You can make gross changes at the start eg. if a rad is dropping virtually nothing then go from fully open (say 8/9 turns) to half shut (4T) in one loop, if it's still dropping next to nothing then half the opening again on the next loop (2T) and fine tune once you start getting meaningful drops. As Phil suggests in his guide I note lockshield open-ness as number of quarter turns from closed as this avoids the use of fractions.

The reason for the tape has been well described here before and yes it is necessary. Your tape is in the wrong place though, hottest place is at the top of the panel closest to the inlet end, coldest place is at the bottom of the panel closest to the outlet end, this is where the patches should be. If you need help determining which is flow and return then read the guide at the link again. A patch of tape about 30mm square is fine (it doesn't need to be black, I use white) and to keep the measurement target small you can just use the IR thermometer in contact with the target (most have a window aperture about 20mm dia).

There is no such thing as a 'required' drop (read the process and philosophy described on the above link again), the system drop is a function of boiler output and total rad loading and unless you have designed it so the rads can sink all the heat that the boiler can produce then the overall return temperature will be higher than the ideal and you wont get the magic 20, 15 or 10 degree drop that you hope for. Throttling the system to compensate for this and force a lower return temperature is not the way to do it, you'll just have to put up with the lower drop.

Have another go and post back if you still have probs.

[John Rumm]

You should be able to get a reading off the painted rad close to the valve position. You won't get a sensible[1] reading of the unpainted pipe or valve metalwork though, without putting a patch of tape on them first and placing the thermometer close to the patch.

[1] The emissivity of the surface will be notably different from that assumed by the thermometer. Some (possibly yours) have the ability to adjust the assumed figure, but its quite a faff getting the number "right" for any given material. Its generally easier to go with the default value that works for most non metalic surfaces and use a bit of tape as a target.

[fred]

In article , fred writes

What I should have said was that the target is the same drop across all rads rather than some target numerical drop. V small rads will rarely achieve full drop though and you just have to live with that/those one(s) being different.

[Percy]

On Wed, 10 Oct 2012, "fred" writ:

My thanks for a thorough response and I can see the sense of what you are saying. I have read so many suggested methods which, in part, seemed often to contradict each other but, the FAQ you pointed me to is easy to understand and makes sense.

This was one method I just couldn't understand the sense of:

'll start again and report back if I don't succeed this time.

[Brian Gaff]

Nothing to do with the problem but the title of this thread had me wondering if it was. Overbalancing radiators, IE they were in a pile and over balanced or You were a novelty act and wre trying to balance radiators, where I wondered? or what in fact you did mean that they were all doing their own thing and casing problems.

Briabn

[fred]

In article , Percy writes

You're welcome and good luck!

I'm qualified to comment on the excessive throttling fault as it's the same mistake I made when balancing my first system many years ago ;-). The writer of that article, Phil, was kind enough to spot my error and correct me, no problems since. The system will run loads better when it is done properly, particularly when the house is up to temp and TRVs start shutting down.

system diagram. I'll not give it further time for a full critique but here's the diagram in case anyone wants a laff:

I'll start again and report back if I don't succeed this time.

Drop us a line anyway, it's good to hear if advice works and how you solved any problems you encountered.

[Percy]

On Wed, 10 Oct 2012, "Brian Gaff" writ:

Brian, I really think you should get out more :o)

[meow2222]

People have explained. Just one point: balancing rad temp drops isn't quite what's wanted, the goal is to balance room temps as near as possible before employing the TRVs. The 2 thing aren't quite the same because a) rad area to room area varies b) wall area to room area varies, affecting heat loss per floor area c) other factors also effect the wanted heat output per room, such as desired temp, exterior wall finishes, sunlight etc

NT

[Phil Addison]

what's wanted, the

The 2 thing aren't

temp, exterior

And the procedure to achieve this is what?

Sorry, but I think balancing radiator temperature drops IS what's wanted. That way each radiator will achieve its design output or as near as the plumbing will permit. The TRV's take up the inevitable mismatch between (a) what the designer had in mind, and (b) what temperature the end requires from day to day.

Of course should you happen to have a grossly oversized radiator it is sensible to throttle its LSV back thus increasing its temp drop and (counter-intuitively) reducing its heat output. This will free up some extra heat to go to other radiators, at least one of which is invariably under-sized.

The rule to remember is that throttling back an LSV reduces the water speed through the radiator, and consequently the water cools more and the outlet temperature becomes lower. In other words the temperature drop increases and the average radiator temperature is lower so less heat is given out to the room.

Perhaps this is what you're getting at? Is it not clear in the FAQ, it's ages since I've read it?

-- Phil Addison

[Percy]

On Wed, 10 Oct 2012, "fred" writ:

As requested, a follow up on progress so far. Starting from scratch as per FAQ, all the upstairs and one downstairs radiator were at least 10C more than the other downstairs radiators. Fearing a long drawn out, running around like a blue a***d fly event, I opted to take a simple measurement of the temperature at the top centre of each radiator as they warmed up, repeated over a number of cycles from cold. I was soon able to bring them all within a few degrees of each other and now find that the radiators that were hottest are within a turn or less of off. The colder radiators are still turned fully on.

None of the drops across the radiators are anywhere near the much quoted

11C but they are all warming up evenly and seem to be within a degree or two of each other when hot. I shall now play about with refining the settings but I am a much happier newbie than I was.

Perhaps the quick and simple method I employed to get to a starting point may be worth mentioning in the FAQ.

Thanks again!

[Roger Mills]

Are the drops more or less than 11 degrees? If less, can you run the pump on a lower setting? This should increase the drop.

I shall now play about with refining the

An even quicker and dirtier method is to go round feeling the rads with a calibrated hand, and to throttle back the hottest ones. Getting exactly the same drop isn't that important - what really matters is making sure that one or two rads are not hogging all the flow to the detriment of the others. As long as they *all* get hot at about the same rate, that's good enough.

[Phil Addison]

It is quite normal to finish with some LSVs nearly off, and the previously coldest rad wide open.

It may well be much quoted, but not by the FAQ. That states "Similarly the oft quoted 11 deg C drop across the radiators is really nothing to do with the radiators! It is the temperature gradient the *boiler* is designed to produce when it is going flat out *and* the pump is producing the specified flow rate. So if you have less than 11C drop it just means that either you have too high a water circulation rate or/and the radiators are not large enough to extract the full rating from the boiler."

As Roger says, if you are getting a lot less than 11 degrees on ALL of them you could benefit by slowing the pump.

That's what we aim for!

Yes the FAQ does need expanding there. Roger has the right idea, starting from cold as soon as the boiler starts to fire, go round and find the ones starting to get hot and throttle them back till the cold ones start receiving hot water. Of course if you already know you have some cold ones, make sure their LSVs are wide open before you start. In a properly balanced system the will all start heating up when the boler fires. Sometimes this is impossible which means there is a design fault in the sytem, usually a too long run of too small a bore pipework.

The FAQ is really written for a new or unknown system.

-- Phil Addison

[John Rumm]

If you want, dump the text from the FAQ into a wiki article and tweak it about, and I will knobble the original FAQ page and point it at the wiki version when you are happy. Or failing that, add some supporting notes on the wiki and I can link them from the end of the FAQ

[Phil Addison]

I'm pushed for time but by all means copy the balancing faq across to the wiki if you care to. Perhaps insert something along lines of "The procedure below is for a new or unknown system, but if your system is 'sort-of' working you might save some time by starting with the existing settings. Then starting from cold, as soon as the boiler starts to fire, go round and find the rads starting to get hot and throttle them back till the cold ones start receiving hot water as well. Of course if you already know you have some cold ones, make sure their LSVs are wide open before you start. In a properly balanced system all rads should start heating up together when the boiler fires.

Sometimes it isn't possible to achieve a balance which means there is a design fault in the sytem, usually one (or more) rad fed with too long a run and/or insufficient pipe bore."

-- Phil Addison

[John Rumm]

if you care

My memory must be going... I went to copy the content, and then found out that I had already done it in August!

starting with

to fire, go

some cold ones,

Added...

[Phil Addison]

Copy of what? ;)

-- Phil Addison

[John Rumm]

I can't remember...

who am I anyway?