Sizing expansion vessels

[Andrew Mawson]

What is the rule of thumb for sizing an expansion vessel for a sealed heating system?

My application is not a domestic heating system but rather an unvented cooling system cooling a large industrial machine, but the principal is the same. The 'boiler' equates to the machine being cooled. the 'radiators' are a fan cooled bank of tubing and the water based coolant is circulated by a pump.

Given the system volume = V and coefficient of expansion of water per degree is E and temperature rise is deltaT presumably :

(V(orig) x E x deltaT) - V(orig) = the max increase in volume

But what safety margins are applied in domestic heating systems? Do plumbers really calculate system volume ?

AWEM

[Ed Sirett]

You should be looking for around 10% of total system volume. That's about 4 times the expected expansion of water from 25-75C.

Allow 10 litres per rad + say 10 litres for a modern boiler.

A 12 litre standard expansion vessel will do 10 rads.

[Roger Mills]

I assume that your system is pressurised, so similar considerations will apply.

With a domestic CH system, the charge pressure (air) in the expansion vessel is typically 0.7 bar with the water system unpressurised. The water system is then pressurised to about 1 bar, cold - which partially compresses the air in the expansion vessel, using up some of its volume. Then the system gets hot - and the volume of water increases by 2 or 3%. The trick is to have sufficient volume in the expansion vessel such that it can accommodate the water expansion without increasing the air pressure (and hence the system pressure) by more than (say) 1 bar.

So if you work out what volume change you need to accommodate, and what pressure rise is acceptable - and apply Boyle's or Charles' Law (forget which!) - you should be able to work out how big the expansion vessel needs to be.

[Guy King]

The message from Ed Sirett contains these words:

Eh? 10 litres per rad and 10 for the boiler is 110 litres, for a 10 rad system, surely?

[Roger Mills]

The *water* volume is about 110 litres. The capacity of the *expansion vessel* needs to be about 10% of this - which is 11 litres - so 12 is ok. What's your problem?

[Guy King]

The message from "Roger Mills" contains these words:

Failure to correctly read your post!

[John]

replying to Roger Mills, John wrote: Be careful with Boyle's and Charles Law These relate to ideal gases and do not work for liquids See Flamco website for a good method to calculate required vessel

[John]

replying to Andrew Mawson, John wrote: normal safety margin applied to heating systems is 25%

[Roger Hayter]

Shame that; it would make cooking really exciting if they did.