Just wondering if I really need hydraulic separation on my Vaillant Arotherm+ ASHP system
(id=1083 on HPM).
I believe the Arotherm+ 5kW internal water pump has a residual pressure of 60kPa (Vaillant manual).
My Grundfos post buffer pump is a UPS3 15-15/65 130 which is on constant speed setting III
which is the maximum and also the factory default - maybe because the installer didn’t think about it ?
The ASHP consists of short run of 28mm copper main flow/return to the buffer followed by 28mm copper joining onto the existing 22mm plastic push fit pipework in the loft running along the centre line of the house. The 8 radiators branch off in pairs on 10mm plastic microbore drops.
Calculated heat loss is 4.6kW@ -3C, 21C inside.
The flow rate of the Arotherm is 860l/hour (0.86m^3/hour) when running under all heat loads (it drops to 680l/hour when cycled off I think).
According to the Grunfos manual (curves attached), the pump should also be providing a head of about 60kPa under all reasonable flow rates. The Grundfos pump should be able to do this up to 1100l/hour (top line on the Grundfos performance curve) ?
I cant really get my head around all the pressure loss stuff, but it seems to me that I am using about 1.2kWh/day running an extra pump for nothing.
Maybe the installer added a buffer pump “to make sure” their modest performance guarantee of SCOP 3.77 was always achieved and I would be better off using the tank as a volumiser ?
I can’t speak about your Vaillant, but I was in a similar position with my LG Therma V. It’s a 12kW unit, which I’m pretty sure is oversized, so I’ve run it all this last season in “quiet mode”, which has had it behaving nicely.
The installers put a 50L buffer in front of the underfloor system and once I’d got it working properly (they didn’t do a great job at all), it was very mediocre in terms of COP for the first year.
Last autumn I decided to convert the buffer to a volumiser (on the flow side) and had the same question about whether I needed one or more pumps after the volumiser and whether they would clash. I looked at the figures and decided the head in metres was going to be close and gave it a go. No calculations, just tried it. And it works a treat. So now I have the pump in the ASHP doing everything, with no zone valves or secondary pumps at all - just flow tweaked to balance the individual zones. The system runs 24x7 with a set point between about 25 and 34 degrees and it’s much, much more efficient.
But what you want to know is how does my pump compare with yours? The pump in my ASHP is a Grundfos with a rated flow rate of 34.5 LMP (2.1 m3/h) and a serviceable head of 8.9m (from the LG installer manual). It’s sited on the ground outside and immediately has to lift the water about 5m to loft height (28mm pipe). Then it circulates down to two underfloor systems (ground and first floor) in smaller pipes. Total floor area in each system is something like 50m2.
If you can make a minimal (reversible?) change to run like this, then it might be worth trying.
I think the issue with my install is that the surveyor never inspected the pipework in the loft and just made a guess and whacked in a buffer to bullet proof the guaranteed SCOP. One of the big energy companies did the system so I had to deal with multiple people and communication chains.
The pipe runs are extremely straightforward in my house which is a single storey bungalow. The existing flow/return from the previous gas boiler runs 15m from the front to the back of the house in the loft. Rads come off in pairs, left and right, as the rooms are either side of the centre line. All new K2 rads. So even though there are plastic pipes, the maximum pressure needed is quite modest. The internal pump in the previous combi gas boiler seem to manage ok - bigger DT but also smaller rads.
My reasoning is that since the Grundfos post buffer pump and the internal ASHP pump are more or less equal then I might get away with just the one pump.
The question is I suppose is how much it would cost and if it could be done reversibly . Another question would be if anyone would want to do it.
Hi Nick. Your system and house are quite similar to mine. I have a 5kW Samsung Gen 6 (No internal water pump HPM #60). Its a bungalow and my 28mm copper pipes attach to the original 22mm copper heating pipes and then onto 22mm plastic (approx 15m) with 15mm plastic tails to my radiators). My heat loss looks to be a bit bigger than yours.
I have no hydraulic separation or volumiser and mine runs very well with a single Grundfos 25-75 130. The pump is controlled by the heat pump itself using PWM.
Not really engaged with your figures properly, but I would generally think that there isnt a lot of saving by removing a pump.. i.e. a single pump would need to be on a higher setting (more pressure head) than two on a lower setting (the same quantity of water is circulated weather 1 or 2 pumps on the system).
That aside, there could be more COP drop with a buffer due to mixing.
These are tempeure measurements from a system. To get 40c to rads, the HP flow has to be 44c
This probably drops COP by around 12%
I was assuming that both my pumps are actually on maximum flow. The internal pump shows 860l/hour on the controller. Maybe that is a Vaillant thing as 860l/hour is the maximum flow rate that the pump is capable of according to the Arotherm+5kW documentation.
I believe it is possible to change this in the settings, but most of the Arotherm+5kW on HPM seem to show same 860l/hour (14.3l/min).
The secondary Grundfos pump has also been set to maximum constant flow (setting III) - not sure this is by design or by default.
Vaillant’s COP figures obviously measure the values in the outside unit and take no account of additional electrical usage of pump and controls or any heat loss in pipework.
So for Feb - Vaillant app gives average COP as 4.30. If the additional electrical energy is added in, this drops to a COP of 3.78. I think the secondary pump uses about 50Watts.
My own not that accurate DIY heat monitor setup also gives an average COP of 3.85 for the same period. I measure the flow and return temperatures next to the buffer ports.
Yes maybe an efficiency drop due to the buffer. It is the short stubby 30litre stackable one from Vaillant with all of the ports at the same level so mixing inevitable unless these is something clever inside it.
Turning it into a 2 port volumiser might help, but I thought that would mean removing the second pump.
I made another dual DS18B20 based wireless sensor a while ago to install on the heating side of the buffer, but there is no room in our tiny tank cupboard to seat it well, so it will have to go on the pipes in the loft. At some stage I have to fix some of the loft insulation moved around by the installers and I will try to add the sensors at that point, and then I may be able to judge the effect of buffer tank mixing as you mentioned. The weather is warming up so I guess any data may have to wait until the autumn now.
