Just wondering what kind of hysteresis blending valves operate to, presume not as tight control as say shower thermostatic valves etc. This will have a big effect on how much water is recirculated by the UFH TMV, presumably none as the TMV is coming up to temperature from cold so all the UFH water is supplied and then boosted by the manifold pump until almost up to the required UFH temperature, has anyone watched this temperature from cold and noticed any over or under shooting or any changes in the manifold flow rates?.
The accuracy of manual and electronic mixing valves is pretty good. There may be a slight hysteresis but I can’t imagine much at all. Remember UFH manifolds should have a thermal cut off switch whereby if the flow temperature to the manifold is excessive (due to a failed valve etc) then power is cut to the pump and no flow is permitted through the loops. This is to protect any screeds, slabs etc and any bare foot homeowners.
Let’s take a working example. Say we have an exact 20kw heat loss at design max load. Say 10kw for UFH and 10kw for space heating. Quite large loads but I’m using it as an example.
Using mad flow rate you will see a 20kw boiler aiming for a ΔT of 20°c will have a flow rate of roughly 14.285 Lpm. Now the UFH which requires roughly three times the flow of a boiler or emitter targeting a ΔT of 20°c has a flow rate of 20.408 Lpm at a ΔT of 7°c across the loops. From those figures you should already see a potential problem arising.
Now let’s say the UFH was off overnight and the slab cooled to 10°c. When the UFH first comes on in the morning the return water to the boiler will be at, or very near to 10°c. Any pipe work above ground might have some negligible temperature difference. If you know the volume of the boiler heat exchanger and the flow rate through the boiler you can calculate how long water entering the boiler takes to leave with a temperature increase of 20°c. Say a 5 litre exchanger and a flow rate of 14.285 Lpm converted to alps = 0.238 Lps. 5/0.238 = 21.008 seconds for water entering the boiler to leave at 20°c higher.
So after 21 seconds we have a new flow temperature of roughly 30°c. Again if you know the volume of pipe work and flow rates you can calculate how long that water takes to reach the manifold. So the water arrives at the blending valve at 30°c but the target temperature leaving the valve will be around 45°c. From here you can see all the flow from the boiler will be pulled by the UFH pump. Now as the slab temperature increases over time so shall the return temperature and in time the boiler flow temperature will rise and the blending valve will need less and less water from the boiler flow. Say the boiler flow is at 65°c and the UFH return 37°c after some time. Using mass flow you can calculate that the UFH will pull around 5.1 Lpm from the boiler, leaving roughly 9.18 Lpm spare if you will. This 9.18 Lpm will be circulating around the space heating, or cylinder, or both, which if designed properly is enough to make the system function.
From the above you should clearly be able to see that if you let a slab go cold overtime then when you do bring the UFH on again it will initially steal all the flow until the slab and return temperature start to rise.
This is exactly why myself and EvilDrPorkChop have stated that UFH should be on constantly. They are designed to be left running to provide a nice even room temperature throughout the day, not switch on and off when you feel like it.