Balancing rads on oversized boiler with varying flow temperature

[Ric2013]

I wonder what the old timers have to say on this issue.

Whenever I try to balance radiators heated by old cast-iron boilers, I'm trying to attain the mythical 11°C across the radiator.

In practice, this seems to be a search for the holy grail because the boiler is invariably running at below full capacity (or has been improperly range-rated) and so is cycling on and off and with a lot of hysterisis meaning the actual temperature of the flow can swing by as much as 10 degrees.

The recent result is that, while I have been 'able' to balance such a system in that each radiator has a maximum of 11°C drop across it, the temperature drop will, at times, drop to zero - and then the boiler refires and re-establishes the 'correct' temperature drop. The radiator is always hot, but sometimes its flow temperature ends up being no higher than the return because the boiler has had to stop firing.

I use one infra-red thermometer. I can get two measurements about a second apart, and I think that is probably as good as simultaneous monitoring of the flow and return of a radiator.

While occupants are happy, and all emitters are getting warm, I do wonder if there is a better way of doing this, but short of having an assistant monitor the flow temperature at the boiler and try to - somehow - manually control the boiler (perhaps by using the boiler thermostat to switch the boiler into firing and idle modes), I can't think of anything. I've worked with a plumber who claims he can get perfect balancing using his hands, but I've never actually seen him do it, and I wonder if what he actually meant was that hands were as precise as you need to be?

And I wonder whether the old timers will say there is any point in perfect balancing, or if I'm worrying about nothing. Does it make a difference to boiler life or efficiency if it isn't spot on?

Thanks in advance for any helpful comment.

 

[JohnD]

surely your objective is not so much the 11C drop, but the relatively comfortable temperature in each of the rooms, with none being over-heated or under-heated? Most importantly, not having any radiators which are cold because some others are taking all the flow? The bare hand will enable you to identify a rad with insufficient, or excessive flow.

Since the 1950's, we now have the advantage of TRVs which give a bit more fine-tuning, once you have balanced them near enough.

 

[Ric2013]

Cheers John. You've hit the nail on the head: what is the 11° all about?

I'd argue that the comfortable temperature of rooms is dictated by radiator sizing and the use of TRVs, but that balancing the radiators is the separate issue of ensuring they receive the exact flow they require to be able to give their stated output with TRV fully open. Then the TRVs partially close because the radiators are often oversized for the rooms (or the system has been designed to keep the room at a temperature higher than that actually desired) and so much for 11 degrees.

My concern is that the 11 degrees (or 20 degrees on a more modern system) is helpful to the boiler to allow the boiler to keep the right temperature (something it's failing to do in this case) and to ensure the boiler runs efficiently (or, in the case of this sort of dinosaur, as efficiently as it can!).

 

[D_Hailsham]

what is the 11° all about?

It goes back to the days when boilers with cast iron hexes were the norm and temperatures were measured in Fahrenheit. Boiler manufacturers decreed that the optimum temperature drop across the boiler was 20°F, so systems were designed for this drop. Along comes the new upstart Centigrade (later renamed Celsius). A quick bit of maths and 20°F becomes =11.111 (ad inf)°C, which was quickly rounded to 11°C.

The purpose of balancing is to ensure that each rad receives the correct amount of heat. That's OK if the rads are correctly sized for each room's heat-loss; but what do you do if they are not? If a rad is undersized the only solution to to replace the rad; if oversized a TRV will help, or you can set the drop across that rad to a higher temperature which reduces the output. A 1kw rad (75/65) will give 830W at 75/55.

Another problem is: what do you do if you can't achieve the magic 11°C (20°C on modern condensing boilers)? The answer is: it doesn't matter. Just balance to whatever temperature works; provided it's not way out. The main reason why it's impossible to achieve the desired temperature drop is that the pump cannot be adjusted to give the correct flow at the actual pressure drop across the system, i.e the required working point on the pump graph is not on one of the three curves (3 Fixed speed pumps). See the graph below. The red dot circled in green is the desired working point (0.7m³/hr @1.8m head) the yellow dot circled in red is the actual working point (0.8m³/hr @ 2.355m head). The increased flow rate means the differential will be lower than desired; a desired 20C differential will become 17.5C.

 

[Ric2013]

Thanks D. Sounds like I'm trying to be a perfectionist for nothing then.

Sounds like consensus is that provided there are no flow issues to the radiators with the greatest head loss, all is well then.

In this case, the pump is not the problem - it is that the boiler flow temperature fluctuates as it is oversized. Range rating it might help, but not my department.

Actually, I was previously having problems getting the balancing right (even though I'd accepted a lesser temperature drop) as the system has quite small radiators and so even a UPS 15-50 pump was serious overkill after I rationalised the ridiculous pipe runs (caused by various house layouts as the house has been extended) and removed the radiators in the porch(!) and the loft(!). Since the pump was replaced with a UPS2 which runs at much closer to the ideal pressure for the system, balancing suddenly became very much easier, but obviously the boiler still cycles.

Thanks for the explanation - I had read of de-rating a radiator (in the same edition of the Reader's Digest DIY Manual that states that a gutter should, ideally, have a fall of 1 in 40) by reducing the flow rate, but wasn't aware that this is or was ever industry practice - after all, now we have TRVs, I doubt it is often done this way. It is good to know the background of WHY we do things the way we do.

 

[hard-work]

The 11C, 80C flow and 59C return, is to prevent condensation occurring in the non-condensing boilers, giving a high temperature return. Condensation can be a big problem with cast-iron boilers. If boiler cycling is a problem, then you can:

1) Fit a timer to give a longer dwell before re-firing the burner.
2) Fit a buffer vessel in the CH line, could be a thermal store. The buffer or thermal store do not care how big the boiler is. They will cut the boiler out when the buffer is up to temperature.

No. 1) above is the easiest and cheapest and the cost of No. 2 is probably more than a proper quality condensing boiler that will drop your gas bills substantially.

But get the system balanced near enough, to start. If the boiler can be range rated get it done.

 

[D_Hailsham]

the pump was replaced with a UPS2 which runs at much closer to the ideal pressure for the system

Is the pump set to a fixed speed (sold green light) or proportional pressure (flashing light)?

I had a UPS2 fitted a few month ago. It is set to a fixed speed as I'm a bit dubious about the proportional pressure settings - the curves for prop pressure do not make sense.

 

[Steelmasons]

The 11C, 80C flow and 59C return, is to prevent condensation occurring in the non-condensing boilers, giving a high temperature return. Condensation can be a big problem with cast-iron boilers. If boiler cycling is a problem, then you can:

1) Fit a timer to give a longer dwell before re-firing the burner.
2) Fit a buffer vessel in the CH line, could be a thermal store. The buffer or thermal store do not care how big the boiler is. They will cut the boiler out when the buffer is up to temperature.

No. 1) above is the easiest and cheapest and the cost of No. 2 is probably more than a proper quality condensing boiler that will drop your gas bills substantially.

But get the system balanced near enough, to start. If the boiler can be range rated get it done.

What a load of tosh...

 

[hard-work]

You are in line for another spanking.

 

[Doggit]

I have never attempted to set the flow across a radiator, and like john, balance the flow into each rad so that they all the rads/rooms come up to temperature at about the same time. And although it's industry standard to fit TRVs (and I do fit them anyway) I'm not convinced that they work that well, and I'm looking forward to trying the new faster reacting acohol based TRVs to see how they do. Getting an 11degree drop across the rads, doesn't necessarliy mean that you'll get the same difference at the boiler.

 

[hard-work]

What make of TRV?

 

[Doggit]

I normally use Drayton, but have tried different ones over the years. Have you any recommendations. I've got a remote type to install soon, and it'll be interesting to see how that works - the rads got a cover on it.

 

[Nige F]

I wonder what the old timers have to say on this issue.


I wonder if what he actually meant was that hands were as precise as you need to be?



.

Worked for me in a 30 bedroom wing of an hotel, the boiler room was about 100 yds away. Precise enough to make a difference afterwards

 

[Ric2013]

Is the pump set to a fixed speed (sold green light) or proportional pressure (flashing light)?

I had a UPS2 fitted a few month ago. It is set to a fixed speed as I'm a bit dubious about the proportional pressure settings - the curves for prop pressure do not make sense.

In what way do they not make sense?

The explanation I had from Grundfos was that since the primaries were always a part of the circuit and the head loss through the primaries would increase with flow, it therefore makes sense to INCREASE the head as the flow increases. So as radiator/zone valves are opened, the pump runs faster to keep the pressure constant, plus a bit more to make up for the increased loss of head through the primaries.

In practice, this seems to work well in my experience, and the pump is set to one of the three proportional curves (flashing).

 

[Ric2013]

The 11C, 80C flow and 59C return, is to prevent condensation occurring in the non-condensing boilers, giving a high temperature return. Condensation can be a big problem with cast-iron boilers. If boiler cycling is a problem, then you can:

1) Fit a timer to give a longer dwell before re-firing the burner.
2) Fit a buffer vessel in the CH line, could be a thermal store. The buffer or thermal store do not care how big the boiler is. They will cut the boiler out when the buffer is up to temperature.

No. 1) above is the easiest and cheapest and the cost of No. 2 is probably more than a proper quality condensing boiler that will drop your gas bills substantially.

But get the system balanced near enough, to start. If the boiler can be range rated get it done.

What a load of tosh...

Steelmasons - why is this tosh?

In reply to hard-work: the system is balanced, subject to debate as in my original post, but consensus seems to be that it is balanced.

Funny that you should mention option 2: a thermal store is actually part of a long-term plan that involves a woodburning stove, and possibly solar thermal, but not the short-term plan.

Not sure what you mean by option 1. The system already has a programmer, if that's what you're on about.

Modern boiler looks good on paper for an average house and an average family, but since the gas bills based on accurate gas readings for this house as it is actually used come to less than £300 a year, it would take a VERY long time to achieve financial payback for a new boiler, so not really an option in this case.