Condensing boiler radiator sizing - CHIC

[Tipper]

Hi, just a quickie, I have an old software programme, about 6 years old, the 'CHIC' heating calculator which I used for radiator sizing before I retired. I now need to use it for my own new condensing boiler heating installations in my flat conversion project and assume the rad sizing will be wrong.

Can anyone advise the %age oversizing of radiators needed to run at condensing flow and return temperatures?

Maybe I can get an upgrade of the CHIC prog?

 

[spudkey]

Look at a rad manufacturers web sites they will have the correction factors for different MWT.

http://www.stelrad.com/UK/docs/compact.pdf

 

[bengasman]

As a rule of thumb, 20% bigger will give you a good result. Err on the side of caution, the bigger the rad, the lower you can set the boilerstat, the more efficient your boiler will be, and the faster your response time.
Apart from taking up space, and costing a few quid more to buy, there are no downsides to oversized rads, but several "problems" if they are too small.

 

[mysteryman]

Put in weather compensation to keep the boiler condensing for as much of the time as possible. There is not much real need for oversizing, as you need the full output only in the bitterest weather. Spend the extra money on improving house insulation.

 

[ianniann]

Put in weather compensation to keep the boiler condensing for as much of the time as possible. There is not much real need for oversizing, as you need the full output only in the bitterest weather.

You contradict yourself. An oversized set of radiators can be operated at a lower flow temperature all the time, hence allowing the boiler to operate more efficiently. Also rooms can be returned from a setback temperature more quickly without using very high flow temperatures, hence the heating doesn't have to come on at midnight just to get warm by 8am Perhaps you are struck in the mindset that "condensing" is either on or off?

 

[D_Hailsham]

Can anyone advise the %age oversizing of radiators needed to run at condensing flow and return temperatures?

It's only the return temperature you need to worry about. Ideally this should be no more than 55°C.

If your boiler works best with a 10°C differential (65°-55°C), the rads need to be oversized by 35%; and if the boiler runs with a 20°C differential (75°C-55°C), the rads need to be oversized by 20%.

 

[ianniann]

A return temperature of 55C is not by any means optimal. I often see this idea that somehow 60C is no good and 55C is "condensing" when actually there is very little difference. The boiler starts to condense at a little under 60C and progressively condenses more and more as the return temperature gets lower, hence getting more energy from the condensing and becoming more efficient. So you aim for the lowest return temperature that is practical in your situation. Most boiler designs max out at a return temperature of around 30C, but in practice you'll never get this with conventional radiators, so just go for the best you can.

Virtually all new condensing boilers are sized for a drop of 20C between the outgoing and return temperatures. The most important thing affected by this is the pump, which can be less powerful when working with a bigger temperature drop. However, the 20C drop won't appear by magic, you have to over-size the radiators to achieve it. Also remember that the temperature differential will go down when some of the TRVs close. And also that it is more difficult to achieve a 20C differential at a low flow temperature, so while it may be easy to drop from 80C to 60C, you'd have to massively over-size the radiators to drop from 50C to 30C.

 

[Agile]

If you look at a current rad list you will see that they are already specified at a flow temp of 70° rather than the old 80°.

There are a few anomalies though and I think the return is taken as 60° and the connection as TBSS.

Tony

 

[D_Hailsham]

If you look at a current rad list you will see that they are already specified at a flow temp of 70° rather than the old 80°.

There are a few anomalies though and I think the return is taken as 60° and the connection as TBSS.

Nearly right.

BS EN 442 specifies 75°C Flow, 65°C Return and 20°C Room temperature

The specified connection is, as you say, TBSS.

 

[Tipper]

Well it took a while but I got the answers in the end!

Thanks guys.

 

[Agile]

Its not quite so easy.

The ideal operating conditions for most current boilers is a flow of 70° and a return of 50° giving an average of 60°.

So the rad is 40° above the room rather then the 50° which rads are specced at. At this time of night I am thinking that rads need to be oversized by 25%.

Tony

 

[D_Hailsham]

It's not quite so easy.

The ideal operating conditions for most current boilers is a flow of 70° and a return of 50° giving an average of 60°.

So the rad is 40° above the room rather then the 50° which rads are specced at. At this time of night I am thinking that rads need to be oversized by 25%.

I agree, it's not quite as easy as you think. Unfortunately the maths is not simple as Newton's Law of Cooling rears its ugly head and the relationship is not linear. So instead of the oversize being 25% for temps of 70/50/20, it is actually 38%.

 

[Agile]

Newton's Law of Cooling is primarily associated with a hot object cooling down.

Whereas a radiator is maintained at fixed temperatures. I would therefore expect the heat loss to be mostly fairly proportional to temperature difference.

How do you arrive at 38% ?

Your previous 20% for 75/55/20 seems to have been too low.

Tony

 

[ianniann]

Newton's Law of Cooling rears its ugly head and the relationship is not linear

Newton's Law of Cooling specifically states that the rate of heat loss is proportional to the temperature difference. However, this isn't strictly true in the case of radiators, and a higher temperature more than linearly increases the heat loss (or heat output in the case of a radiator).

As an example, the output for a 60C (above ambient) radiator is approximately twice the output at 35C above room temperature, and 2.5 times the output of a radiator at 30C above. By far the easiest way is to use a table of correction factors, provided in all good radiator sizing guides.

For the same reason of not being strictly proportional, the output of a radiator is not exactly determined by the average of the flow and return temperatures, but this difference is usually small enough to ignore. I've seen tables that give corrections for different combinations of flow and return temperatures and they are just unnecessarily complicated. In any case, different radiators with different flow rates will have a different pattern of temperatures over its surface, so trying to get too precise is pointless.

The ideal operating conditions for most current boilers is a flow of 70° and a return of 50°

Again this idea that there is some "ideal" operating temperature for a modern (condensing I presume) boiler. A typical boiler is 3-4% more efficient at 70/50C than at 80/60C and another 3-4% better at 50/30C. The exact details depend on the individual boiler and also on the modulation level of the boiler. Most boilers will quote the efficiency or nominal output at full output at both 80/60C and 50/30C, and occasionally for other output levels and temperatures. You will see that operating underfloor heating at perhaps 40/30C is more efficient than running radiators at what you call the "ideal" 70/50C.

 

[D_Hailsham]

How do you arrive at 38% ?

Your previous 20% for 75/55/20 seems to have been too low.

The equation I used is the "Radiator Enlargement Ratio" as shown in Formulas published by Danfoss Heating. The same formula is used for sizing radiators on a single pipe systems due to the drop in flow and return temperature as you proceed along the chain from the boiler.

If you look at the Temperature Table on page of the Stelrad Elite Catalogue, you will see that the factors are not linear. (This table shows the change in rad output from norm) You would expect a rad running with a mean difference of 40C to produce 80% of the output of one running with a mean of 50C, but it doesn't. As the table shows, it only produces 74.8%. Similarly for a difference of 25C the output is 40.6%, not the expected 50%. This table assumes that the difference between flow and return temperatures stays constant at 10C.

If a boiler is used which runs best with a 20C differential, the factors are virtually the same, provided the mean rad temperature is the same, e.g 80/60 instead of 75/65.

In the example I gave with a boiler running at 70/50 the mean temperature is 60C and the difference is 40C (60-20, room temperature).

Using the table in the Elite catalogue the factor for a 40C difference is 0.748; i.e a nominal 1000W radiator will give out 748 watts. So if you need 1000 watts, you will have to install a rad which is nominally 1000/0.748 = 1,337 watts., an approx 34% oversize

However if you use the formula I gave the link to, you will find that the enlargement factor is actually 1.378, i.e an approx 38% oversize.

I haven't been able to account for the few percentage points difference.