A few variables there. Firstly it would be interesting to know what type of boiler you have and whether the pump is internal to the boiler or external.
The old gas boilers (and most modern oil boilers) had a fixed rate at which they burnt gas and the burner was on or it was off. The traditional temperature drop was 'around' 11°C, simply because that was 20°F which was a nice round number that was a reasonable compromise between being too high a drop (result being colder radiators and greater thermal stress on the boiler heat exchanger) and too low a drop (the circulation pump would be oversized and noisier, use more electricity etc). The 11°C drop was with the boiler running at maximum temperature (usually 80-82 degrees) on the flow and firing continually, with the rooms up to temperature.
In an ideal hypotherical situation, the old boilers would be running continually and the flow temperature would reach 80°-82°C and then not get any hotter because the heat lost to the rooms by the radiators matched the heat put into the water by the gas being burnt at the boiler. In this situation, balancing is fairly easy.
Modern condensing gas boilers are usually quite happy running at a 20 degree drop, radiators are larger to allow the boiler to run cooler (often 70/50). They may vary their flow temperature to try to limit the drop to a maximum of 20°C. Or, if a new boiler is coupled to an old system, then the radiators may remain at 80/70 or 82/71 to allow for sufficient output. The flow temperature can still be reduced in warmer weather.
The problem with balancing is that boilers often have a heat output that has been set to a rate greater than the total output of all your radiators put together. In this case, it is sometimes hard to get the boiler to fire continually as setting the gas burn rate is a boiler adjustment that isn't really for DIY or even for someone like myself that is not a registered gas installer, so the boiler has to switch the burner on and off to avoid overheating the water. A modern condensing boiler may ramp down the gas burn rate to allow them to run without cycling on and off, but many old boilers will cycle on and off, and many modern condensing boilers may not cope with the load they are being asked to give and will tend to cycle too.
Logically, if you find yourself in this situation, then you are probably better off paying more attention to the return temperature and trying to get that around the same for each radiator. How much cooler than the flow? If it's a condensing boiler that can vary its burn rate 'modulate', then I'd be looking to get it within 20 deg C, whereas if it's an old boiler, I'd be aiming at 11 deg C, and no more than 12. But if your radiators receive their flow at 65 instead of 80, then you might consider that 12 degrees is an excessive drop as the 11 degrees figure was for a flow of 80. If you find the difference is only 7 degrees, say, with a flow of 80 degrees, you may be able to slow the pump down and see if that gives you a better drop, but it doesn't really matter too much and I doubt this will be the case in your circumstances.
If the boiler is switching on and off then the flow temperature may fluctuate. In this case, I try to take the peak flow temperature as being the reference to measure against, but I'm not sure this is universal practice and other plumbers may have different ideas.
On to your 40°C question. I am guessing you are measuring the temperature with the boiler not yet having reached its maximum temperature (because it seems unlikely you would run a boiler this cold). How does the the flow at the boiler outlet compare with the temperature at radiator inlet and vs other radiators? If within a few degrees, then it seems to indicate that the boiler is not yet up to temperature. If, on the other hand, the boiler is getting hot and the radiator is not, there could be a partially blocked pipe (sludge or air) or a fundemental design issue. I have seen additaional radiators piggybacked off other radiators in an existing system with the result that the water starts to cool before it gets to the radiator: there isn't much you can do in that case unless you want to start increasing pipe diameters within the system. The fact that the radiator is running without a temperature drop suggests it is receiving excessive flow and needs throttling back, and you might find that you have a modern boiler and that it is reducing its flow temperature (or even slowing down the pump) to try to compensate for the fact that it sees that no heat has been lost by the water it has sent around the radiator system.
Back to our hypothetical situation from paragraph 3. As soon as you add TRVs, this situation becomes unsustainable. As the temperature outside rises above the -1 to -3°C that the system is probably designed to work at, the heat losses from the house fall and the radiators start to overheat the rooms. The TRVs shut down the radiators, the boiler is running at part load, and the drop across the radiators increases. As such, there is a certain amount of truth in the idea that TRVs will self-balance a system. The problem is when you first switch on the system and find some radiators are taking 45 mins to warm up, or someone turns a TRV to max and starves other radiators. By manually balancing the system, what you are avoiding is that, but the TRVs will usually take over control of the balancing most of the time. So, if the number of variables means that the hypothetical ideal is not sustained, it doesn't really matter because you'll have achieved a good enough balance, even if absolute perfection was not possible.
Does this help?
The old gas boilers (and most modern oil boilers) had a fixed rate at which they burnt gas and the burner was on or it was off. The traditional temperature drop was 'around' 11°C, simply because that was 20°F which was a nice round number that was a reasonable compromise between being too high a drop (result being colder radiators and greater thermal stress on the boiler heat exchanger) and too low a drop (the circulation pump would be oversized and noisier, use more electricity etc). The 11°C drop was with the boiler running at maximum temperature (usually 80-82 degrees) on the flow and firing continually, with the rooms up to temperature.
In an ideal hypotherical situation, the old boilers would be running continually and the flow temperature would reach 80°-82°C and then not get any hotter because the heat lost to the rooms by the radiators matched the heat put into the water by the gas being burnt at the boiler. In this situation, balancing is fairly easy.
Modern condensing gas boilers are usually quite happy running at a 20 degree drop, radiators are larger to allow the boiler to run cooler (often 70/50). They may vary their flow temperature to try to limit the drop to a maximum of 20°C. Or, if a new boiler is coupled to an old system, then the radiators may remain at 80/70 or 82/71 to allow for sufficient output. The flow temperature can still be reduced in warmer weather.
The problem with balancing is that boilers often have a heat output that has been set to a rate greater than the total output of all your radiators put together. In this case, it is sometimes hard to get the boiler to fire continually as setting the gas burn rate is a boiler adjustment that isn't really for DIY or even for someone like myself that is not a registered gas installer, so the boiler has to switch the burner on and off to avoid overheating the water. A modern condensing boiler may ramp down the gas burn rate to allow them to run without cycling on and off, but many old boilers will cycle on and off, and many modern condensing boilers may not cope with the load they are being asked to give and will tend to cycle too.
Logically, if you find yourself in this situation, then you are probably better off paying more attention to the return temperature and trying to get that around the same for each radiator. How much cooler than the flow? If it's a condensing boiler that can vary its burn rate 'modulate', then I'd be looking to get it within 20 deg C, whereas if it's an old boiler, I'd be aiming at 11 deg C, and no more than 12. But if your radiators receive their flow at 65 instead of 80, then you might consider that 12 degrees is an excessive drop as the 11 degrees figure was for a flow of 80. If you find the difference is only 7 degrees, say, with a flow of 80 degrees, you may be able to slow the pump down and see if that gives you a better drop, but it doesn't really matter too much and I doubt this will be the case in your circumstances.
If the boiler is switching on and off then the flow temperature may fluctuate. In this case, I try to take the peak flow temperature as being the reference to measure against, but I'm not sure this is universal practice and other plumbers may have different ideas.
On to your 40°C question. I am guessing you are measuring the temperature with the boiler not yet having reached its maximum temperature (because it seems unlikely you would run a boiler this cold). How does the the flow at the boiler outlet compare with the temperature at radiator inlet and vs other radiators? If within a few degrees, then it seems to indicate that the boiler is not yet up to temperature. If, on the other hand, the boiler is getting hot and the radiator is not, there could be a partially blocked pipe (sludge or air) or a fundemental design issue. I have seen additaional radiators piggybacked off other radiators in an existing system with the result that the water starts to cool before it gets to the radiator: there isn't much you can do in that case unless you want to start increasing pipe diameters within the system. The fact that the radiator is running without a temperature drop suggests it is receiving excessive flow and needs throttling back, and you might find that you have a modern boiler and that it is reducing its flow temperature (or even slowing down the pump) to try to compensate for the fact that it sees that no heat has been lost by the water it has sent around the radiator system.
Back to our hypothetical situation from paragraph 3. As soon as you add TRVs, this situation becomes unsustainable. As the temperature outside rises above the -1 to -3°C that the system is probably designed to work at, the heat losses from the house fall and the radiators start to overheat the rooms. The TRVs shut down the radiators, the boiler is running at part load, and the drop across the radiators increases. As such, there is a certain amount of truth in the idea that TRVs will self-balance a system. The problem is when you first switch on the system and find some radiators are taking 45 mins to warm up, or someone turns a TRV to max and starves other radiators. By manually balancing the system, what you are avoiding is that, but the TRVs will usually take over control of the balancing most of the time. So, if the number of variables means that the hypothetical ideal is not sustained, it doesn't really matter because you'll have achieved a good enough balance, even if absolute perfection was not possible.
Does this help?
