Evohome and balancing

[DorrisMancer]

I raised balancing in an earlier thread but still can't get my head around it. I'm quite prepared to believe that I've missed something or got something wrong - but I can't see what it is. If you think that balancing is worthwhile and possible in an evohome system then please tell me where I'm wrong so I can stop thinking about it.

My view of the basic facts:

1. Lockshield valves (LSVs) in simple systems, without TRVs or multi-speed pumps, were set to balance the flow distribution to all rads. LSVs in condensing systems were also set to ensure the temperature of the return water was low enough to get the boiler condensing.
2. Water in a rad loses heat in direct proportion to the temperature difference between the water and the room
3. Water in a rad loses heat in direct proportion to the time it spends in the rad, and therefore inversely with the flow rate, but the flow also introduces new water
4. LSVs are adjusted with TRV/HS92s fully open therefore TRV/HS92s can only reduce the flow thereafter

It's clearly possible to adjust the LSVs to give a 20degC differential with a fixed flow rate pump, with the TRVs fully open, and with a particular water-room differential temperature.
But, as TRVs close or as the rooms heat up this differential setting will be lost.
With an evohome system the HS92s are closing asynchronously by varying amounts, plus the modulating pump is changing the flow rate, plus the rate of loss of heat is changing as the rooms heat. Unless someone can explain to me how balancing can be maintained in any meaningful way in this situation I suspect that the real reason we adjust the LSVs is just to get a balance of flow when lots of rads are calling for heat.

Please explain what I've missed or got wrong!

[dty]

Plus with OpenTherm the flow rate is constantly changing, further fiddling with dT.

The biggest argument is about when all the rads open up simultaneously - say, first thing in the morning. Although with sensible zoning and optimum start, it's likely that different zones will start up at different times any way.

We have a 25kW boiler and 25 rads in a 4,000 sqft home. I'm almost certain that the boiler is too small, despite being replaced in May this year. However, it works because there's never a demand to heat the entire house from cold simultaneously. Except... when we come back from holiday. It can actually take a day or so before the house feels comfortable. And it's noticable that the rooms furthest from the boiler suffer the most because of poor balancing and therefore flow starvation.

Which is all to say... sometimes you DO want to heat the whole house at once. But mostly you don't.

[DorrisMancer]

That's a refreshing first response, I'd expected to be shot down From the above I assume you've set your LSVs to get heat to the furthest rad, rather than trying to get any particular dT ... or did you set them all wide open?

We're about the same size but very leaky, 32kW W-B (so no opentherm) and too early to say how well the boiler is sized - it's a bit bigger than called for by the calculations.

[HenGus]

I have had Evohome (19 radiators) for nearly 4 years - initially, with an old Ideal Classic boiler and now with an ATag under Opentherm control. My radiators have not been balanced and I haven't noticed any issues with slowly heating radiators etc. Once the zones are within their set target range, Opentherm reduces the TSet temperature. Currently, TSet is 62C and I am getting a return flow of 52C. My boiler output range is 5 to 24kWhs. As I write, even 5kWs is too high for the heat that is needed. Based on gas usage, energy consumption was only 4.36kWhs over the past hour. Would radiator balancing give any tangible improvement - I confess that I do not know.

[filbert]

Our system was installed recently and one of the rads near the boiler had its lock shield opened wide (it was the system drain point and the system had to be drained to add a rad and replace TRVs). This rad then came on blazingly hot and the room got over warm (it isn't one we use a lot and the temp should be around 16 to keep it on tickover). The room next to is had a similar issue, even though we use it and it is set to 20deg, it was getting too warm.

I set the lockshield to around a quarter of a turn and it's much better. I can't say how much this was affecting other rads but it might well have been.

I must admit, I thought balancing wouldn't be needed as the HR92s would do all the work but I think I was wrong. At some point, I intend to buy a thermometer (maybe this one) and balance it - but I can't see a pressing need, as the system is pretty much OK.

[DBMandrake]

I raised balancing in an earlier thread but still can't get my head around it. I'm quite prepared to believe that I've missed something or got something wrong - but I can't see what it is. If you think that balancing is worthwhile and possible in an evohome system then please tell me where I'm wrong so I can stop thinking about it.

Balancing a system is independent of what type of TRV's you use and whether they are smart or dumb. Balancing is performed when all TRV's are fully open.

My view of the basic facts:[*]Lockshield valves (LSVs) in simple systems, without TRVs or multi-speed pumps, were set to balance the flow distribution to all rads. LSVs in condensing systems were also set to ensure the temperature of the return water was low enough to get the boiler condensing.

Not quite right. Lockshield valves are set to balance the flow resistance between different branches of the piping which may have different diameters and lengths, so that the water finds it roughly equally easy to flow to every radiator so that some radiators wont steal most of the flow from others when all the TRV's are open. The radiator with the most flow resistance (usually the furthermost one, but it does depend on diameter of the pipes and how close it is connected to a large diameter trunk) usually has its lockshield valve left fully open and ones that have less flow resistance have theirs partially closed.

It's the speed of the pump that is set to give the correct temperature drop across the radiators. If you just leave the speed on maximum and try to close down every lockshield valve to get your desired temperature drop all you'll succeed in doing is diverting most of the flow via your automatic bypass valve which will have consequences for condensing efficiency. (The ABV shouldn't flow at all unless you are down to 1-2 radiators open)

To get everything right there might be a bit of back and forth between the pump speed and the lockshield adjustments.

[*]Water in a rad loses heat in direct proportion to the temperature difference between the water and the room

Not quite - it loses heat in proportion to the difference in average surface temperature of the radiator and the room. This is important when you consider there is a temperature drop across the radiator. If the feed side is 70 and the return side of the radiator is 50, then 60 is the figure you use to calculate heat output.

[*]Water in a rad loses heat in direct proportion to the time it spends in the rad, and therefore inversely with the flow rate, but the flow also introduces new water

Not quite sure what you're trying to say here. If you're trying to say that a flow rate that is too high will cause less heat to be transferred from the water to the radiator and on to the room then that is not the case, quite the opposite. Reducing the flow rate always reduces the heat output into the room, and for two reasons.

One is that a lower flow rate will cause the water to stay in the radiator longer causing more heat loss from a given quantity of water which means the return side of the radiator will get colder and you will have a greater differential across the radiator, this will reduce the average panel surface temperature and therefore average heat output.

The second reason is that water has a finite specific heat capacity. To transfer heat you must transfer sufficient volume of water per unit time to transfer that heat. If the flow is too slow you limit the amount of heat that can be transferred and you will find not only will the cold return side of the radiator get colder but so will the hotter supply side.

These are after all the mechanism by which a TRV is able to control the heat output into the room.

[*]LSVs are adjusted with TRV/HS92s fully open therefore TRV/HS92s can only reduce the flow thereafter

Yep. Because you're setting the lockshield valves to balance the flow resistance of the different branches. That means all HR92's are starting on a level playing field when fully open and have a similar range of control over their radiator.

It's clearly possible to adjust the LSVs to give a 20degC differential with a fixed flow rate pump, with the TRVs fully open, and with a particular water-room differential temperature.
But, as TRVs close or as the rooms heat up this differential setting will be lost.

Of course it will be "lost" - closing the TRV reduces the flow through the radiator and will cause the differential across the radiator to increase to reduce heat output into the room. That's it's job...

What makes you think that a radiator should have a 20 degree differential at all TRV openings ? It clearly can't and shouldn't. After all the TRV is restricting the flow in the exact same way that the lockshield valve you just adjusted to achieve 20 degrees does, (just on the opposite side of the radiator - but that's immaterial) so why wouldn't it affect the differential ?

You're also assuming that the point of balancing a system is solely to try to get a 20 degree differential across the radiators - that is a bit of a pet peeve of mine because although that's what you'll usually read I believe it is misleading. The real reason to balance is to balance the flow resistance between different branches so that when all TRV's are open the radiators flow equally and that a radiator near the boiler doesn't starve further away radiators for flow. Once you have this balance you can then adjust the pump speed of your still-balanced system to achieve your target temperature drop across most of the radiators.

With an evohome system the HS92s are closing asynchronously by varying amounts, plus the modulating pump is changing the flow rate, plus the rate of loss of heat is changing as the rooms heat. Unless someone can explain to me how balancing can be maintained in any meaningful way in this situation I suspect that the real reason we adjust the LSVs is just to get a balance of flow when lots of rads are calling for heat.

Please explain what I've missed or got wrong!

I think what's confusing you is you are under the impression that balancing a system is to try to achieve a 20 degree drop across radiators at all times. It's not, and it's not even possible nor desirable. It's to balance the flow resistance of each branch. Once each parallel branch has an equal flow resistance (with TRV fully open) each individual HR92 can control its radiator over a roughly equal range of flow rates.

And even when some radiators are closed the ones that are still open will be balanced in relation to each other and in relation to how far their individual TRV's are open.

If a system is badly balanced then certain radiators may not be able to heat up at all when most of the other radiators are fully on, and I have one such radiator in my house, which is the furthermost from the boiler with a long run of microbore. With all the lockshield valves in the house fully open that radiator literally does not heat up when all other radiators are on.

In such a situation a simultaneous start of all radiators (such as manually turning the heating on in a cold house, thus avoiding optimal start or differing schedules) means that room will stay cold until many of the other rooms are up to temperature enough for their TRV's to start closing. Not good.

With my radiators balanced all the radiators in the house warm up at roughly the same rate from a complete cold start including the far away one. There is no room that is starved for heat when the other rooms are heating, and that is why balancing is done.

Does every house need balancing ? Probably not. It depends a lot on the size, length of pipe runs and how they are all connected together. Certain configurations are inherently better balanced.

For example if you had a main 22 or 28mm header which ran around most of the house and then 15mm tails from the header to each radiator that were no more than about a metre long (and about the same length for every radiator) then the system would be pretty well balanced without adjusting anything. All you'd need to do is set the pump speed for the temperature drop.

On the other hand if you have long 15mm runs from the 22/28mm header and especially if they differ in length greatly from one radiator to another or there is any additional branching in those 15mm runs then balancing will be required. If you have 8mm microbore like me with greatly differing length runs and some radiators with 15mm then you definitely need to balance the system.

My approach to checking the balancing of the system is turn the heating off at night (including any schedules/timers) so the house is good and cold in the morning, turn every radiator on at once to a high temperature, then walk around the house feeling each radiator and comparing how quickly they heat up - keep walking around quickly from room to room touching the radiators, and keep doing this until they are all fully hot. Ideally they will all feel about the same temperature as they warm up.

If you have an IR temperature gun you can do this a bit more scientifically - just aim it at the middle of the panel so you can get an average temperature of the panel, although beware that an older radiator with sludge, or a radiator with air pockets that needs bleeding will have cold spots that may lead to a false reading during the early stages of heating before heat starts to flow through the steel work of the panel.

If any are racing ahead or lagging behind the others, you need to balance the system. Overall temperature differential should be achieved by adjusting the pump speed.

[Shadylee]

What about the situation where the pump speed is fixed?

My vokera excel does not have any pump speed adjustment. But I do wonder if it is inverter driven and modulated?

If it's a fixed speed pump surely the only way to drop return temps is to calm the flow through the rads?