Posted by: @iaack

↑

Yes, you should plumb a domestic fan coil radiator differently if you plan to use it for cooling (in addition to or instead of heating)

Here's a link to the Strada brochure.  You'll see that the rads are designed to perform both heating and cooling using the same connections.  Because the emitter is essentially horizontal connections there's no top or bottom and both connections are at the same end.

Clearly forced air circulation is required for cooling and Jaga supply fan units as options to sit on top of the emitters.

Jaga state that light cooling is possible without condensation problems as long as the flow temp isn't too low.  We find 15C is usually fine.  

@downfield Ah, sorry I didn't realise you had fan assisted convected radiators, I read the data sheet and see that these are non condensing so the flow needs to stay slightly higher in cooling but it also says in the capacity table that the cooling kW is 40% of the heating so there is the issue with the cycling on a 14kW ecodan.

Sorry I missed that detail.

Posted by: @ashp-bobba

↑

Not to mention the DT, if the room is 30 and the flow is 17 /19 (M18) = (30-18=12 Rad DT) where your system on heating would have been designed closer to RoomDT 25 on heating assuming target room temp 20 flow (M42.5) 

We have the equivalent of 8x Strada rads which absorb around 650W each on max fan speed at a room temp of 27C with flow @16C and return @18C according to the specification.

We also have 3x 1800mm trench rads (Clima Canal) which absorb about 800W each.

They are designed to do this, so we're not trying to second guess how they do it.

So a total of 5200W + 2400W =7,600 W of cooling which we find works well.

The point of my original post was to see if it could be run more efficiently and it seems that the buffer might be the cause of the frequent cycling.

But that aside I wonder why the ASHP is producing flow at 12C when set at 15C on the controller?

Posted by: @downfield

↑

But that aside I wonder why the ASHP is producing flow at 12C when set at 15C on the controller?

Overshoot due to small system volume, with a control loop that assumes a larger volume?

Just a guess.

I think Mitsubishis cooling capacity tables might include min output, their heating ones certainly do.  It would be worth taking a look to see how this stacks up and whether it helps identify the cause.

@downfield Hmm, with the 12Deg C it may be an overshoot, they do cycle down and overshoot slightly if they cant chase the DT across the plate HX, I would not be surprised at 1 or 2 Deg but 3 seems a slow reduction or even a hard stop at the end. 

So to work out if the buffer is the issue you would need to measure if the capacity that you are running to the buffer matches the outlet to the Strada rads. You could do this by, flow rate calculation to the buffer, PA prop at buffer, flow rate from buffer and the calculate each circuit and check the DT on each rad, there are also flow meters you can add to each circuit if you wished.

The 14kw will still short cycle a bit on 50% capacity but I would have though much lass than you are indicating as they can run at 50%. I think you are correct, the buffer could be the cause and could be unbalanced from the emitters. 

Just as a fun experiment and if the 2 pumps are controllable on speed settings you could slow the HP-BUffer pump by 1 setting and speed up the Buffer to Emitters pump by one setting as I suspect (without knowing much more and an investigation) that the emitters although 7Kw and above are not pulling as fast as the HP is producing. You can read the flow rate of the primary circuit via the ecodan controller, that would give you a starting point.

@jamespa you are so much faster than me, happy to see great minds know the same stuff 🙂

Posted by: @jamespa

↑

I think Mitsubishis cooling capacity tables might include min output, their heating ones certainly do.  It would be worth taking a look to see how this stacks up and whether it helps identify the cause.

Thanks - I will check this.

Posted by: @ashp-bobba

↑

Just as a fun experiment and if the 2 pumps are controllable on speed settings you could slow the HP-BUffer pump by 1 setting and speed up the Buffer to Emitters pump by one setting as I suspect (without knowing much more and an investigation) that the emitters although 7Kw and above are not pulling as fast as the HP is producing. You can read the flow rate of the primary circuit via the ecodan controller, that would give you a starting point.

Thanks - I will try this when the loft temp goes down a bit!

On the buffer tank issue, I use Perplexity more and more as a search engine and I asked it whether a 3 pipe buffer would be a good compromise for both heating and cooling.

It led me to an interesting diagram in an article in a US Plumbing and Mech Eng magazine by John Siegenthaler on this subject dated 2023.

You have to register to search the back copies, which I did and found the attached article, which is very relevant.  I made a pdf and this is attached.

He outlines various options but the three pipe arrangement is about half-way  through the article, and uses 4-way valves to reverse the flow through the buffer.

His logic seems to make sense to me but of course there is the added complexity and the need to control the valves, if not switching manually - which I could, twice a year, but my wife would not want to.

After reading the extensive discussions on the pros and cons of buffer tanks for heating, I was thinking of switching to a volumiser arrangement, with a single pump.

But the 3-pipe set up described in the article seems better for cooling and I can keep the 2 pumps in case I ever need to switch back or try something else.

Grateful for any thoughts on this arrangement...

Posted by: @downfield

↑

On the buffer tank issue, I use Perplexity more and more as a search engine and I asked it whether a 3 pipe buffer would be a good compromise for both heating and cooling.

It led me to an interesting diagram in an article in a US Plumbing and Mech Eng magazine by John Siegenthaler on this subject dated 2023.

You have to register to search the back copies, which I did and found the attached article, which is very relevant.  I made a pdf and this is attached.

250701 9DC Buffer tank options.pdf

He outlines various options but the three pipe arrangement is about half-way  through the article, and uses 4-way valves to reverse the flow through the buffer.

His logic seems to make sense to me but of course there is the added complexity and the need to control the valves, if not switching manually - which I could, twice a year, but my wife would not want to.

After reading the extensive discussions on the pros and cons of buffer tanks for heating, I was thinking of switching to a volumiser arrangement, with a single pump.

But the 3-pipe set up described in the article seems better for cooling and I can keep the 2 pumps in case I ever need to switch back or try something else.

Grateful for any thoughts on this arrangement...

 

I have seen this discussion (not exactly this article) before and the argument for a 3 port buffer over a 4 port buffer does indeed make sense to me.

However nothing in the article addresses the question 'why put a buffer in in the first place'?  Put another way - what problem is it trying to solve that cannot be solved by an arrangement that does not risk mixing (also called distortion)?  Unless there is an answer to this question, I would still say get rid.  Without a doubt buffers can be beneficial for installers by masking problems that would otherwise result in a call out (but should be fixed) and circumventing any responsibility for the pre-existing components, but that's no advantage to householders!

If further evidence is needed, its notable that @mars has now done two podcasts about buffers in a domestic situation, each time inviting those installers who are 'for' and those installers who are 'against' to take part.  On neither occasion would anyone 'for' agree to appear.  That, alone, speaks volumes.

Regarding 'But the 3-pipe set up described in the article seems better for cooling and I can keep the 2 pumps in case I ever need to switch back or try something else.' you could also do this by bypassing the pump with a couple of valves and likewise bypassing the return side of the buffer tank with a couple of valves.  That would enable you to switch back more or less instantaneously. 

@jamespa I am not in principle against removing the buffer - my main concern is system volume.

The Jaga rads hold very little water - less than a litre each and so with the pipework only perhaps 20 litre in total?  I'm guessing.

Wouldn't the 14kW Mitsubishi need more than this?  perhaps converting to a volumiser is the way to go?

Posted by: @downfield

↑

@jamespa I am not in principle against removing the buffer - my main concern is system volume.

The Jaga rads hold very little water - less than a litre each and so with the pipework only perhaps 20 litre in total?  I'm guessing.

Wouldn't the 14kW Mitsubishi need more than this?  perhaps converting to a volumiser is the way to go?

I absolutely wouldn't remove the buffer altogether, I'd convert it to a volumiser.  My suggestion above was intended to convey that sorry for not being clear.

Even that is a bit too small for margin.  Suggest to check the specs 

PS, found the specs here (page B-63 - not exactly where you would expect them to be in the document), 20l or 60l depending on cool vs average climate.  I think we are in the latter but given that defrost occurs a lot here and you are having a problem with cycling Id personally go for the higher figure, which I think your BT converted to a volumiser will just achieve.  More is better.

The same doc also has cooling min capacities.

(Health warning - Im not a heat pump installer, just a homeowner with a background in physics who has been studying this stuff for over 2 years)