Ecodan - adding a zone-specific buffer tank ?

[joth]

Backstory: I have ecodan 8.5kW with 100m2 UFH on the ground floor, plus a small FCU in the loft "just in case" (thread)

 

The FCU is connected as 'zone 1' per this diagram, with a tiny (10L?) low-loss header:

 

 

 

 

Problem is we're using the FCU far more than anticipated, for cooling upstairs, and it's very inefficient as it causes ASHP short cycling (thread).  unfortunately space is tight in the plant room so very difficult to replace the LL header with a bigger (200L) buffer tank.

My current line of thought is to put a large buffer tank in the loft, next to the FCU. Downside is this means one more pump (to loop out of the buffer into FCU, as the existing zone 1 pump would just work to fill the buffer). The UFH won't get the benefit of the buffer, but it doesn't need it as it's in 80mm screed slab which acts as its own buffer / heat store.

 

The other option maybe to put the buffer in the loft, and remove (bypass) the low loss header altogether, so in essence the buffer is fed from the primary circulation pump, and the UFH can if needed (e.g. mismatched flow vs primary pump) back-draw from the buffer if needed by pulling water back down the "up" flow path to the loft. This is.... dodgy? but I think plausible? 

[dpmiller]

why not just put the buffer in series with the FCU? It'll take time to cool (or heat!) the mass of water.

[joth]

4 hours ago, dpmiller said:

why not just put the buffer in series with the FCU? It'll take time to cool (or heat!) the mass of water.

Yes I think that would be my plan A (as in "My current line of thought is to put a large buffer tank in the loft, next to the FCU." - sorry I see I didn't spell out the exact schematic much)

 

I assume the buffer will give preferential feed-through to the FCU load, thus can get immediate cooling/heating even before the buffer has charged up, but, does mean a 4th pump (and a bunch more control logic - fine, I can do that in Loxone).

 

I'll try and scribble some schematic options

[dpmiller]

put the buffer on the return.

[joth]

15 hours ago, dpmiller said:

put the buffer on the return.

Love it. 

So only 2 ports needed. I'll get figuring out the capacity to get the cycle time up to, what is good, 30mims minimum?

 

I would go massive with it, the only downside is the extra glycol then needed to fill the system. 

[joth]

A very quick looks, seems like a direct UVC is about half the price of a similar sized buffer tank, and gives me an immersion (and stat?) pocket which would provide a backup heating option incase of ashp failure

Is there any practical difference?

The system already had a pressure release valve and expansion vessel. Not sure if using a UVC like this would mean another is needed, or if it triggers G3 paperwork needs. 

 

https://www.heatershop.co.uk/gledhill-stainless-es-unvented-90-litre-cylinder-direct-connection

 

Not sure if I'll do this this season or wait until after winter now. Good to have a plan in place anyway ?

[Dan F]

On 30/07/2021 at 13:11, joth said:

replace the LL header with a bigger (200L) buffer tank.

 

I've seen buffer size discussed multiples times on the forum, but did anyone ever come up with any "buffer maths" to calculate buffer sizes based on flow temps and heating circuit characteristics?

[Nickfromwales]

1 hour ago, Dan F said:

 

I've seen buffer size discussed multiples times on the forum, but did anyone ever come up with any "buffer maths" to calculate buffer sizes based on flow temps and heating circuit characteristics?

It'll be reasonably unique to each instance. System volume, pipe centres, actuators vs no actuators and so on.

 

100L should more than suffice as a one-size-fits-all solution tbh, but it's then down to whether it's deployed as a 'thermal store' or an 'energy buffer'.