Flow rate needed

[jamesdiyer]

1 hour ago, dpmiller said:

all those elbows on the back if the unit will be adding massively to restriction

Can I calculate how much? So I can say to the installer, if we improve this I think flow will improv by X? 

[MikeSharp01]

38 minutes ago, jamesdiyer said:

Can I calculate how much? So I can say to the installer, if we improve this I think flow will improv by X? 

In theory yes, but might a couple of large diameter flexis not make all the transitions smoothly and so remove the need for the 90 deg angles everywhere.

 

[jamesdiyer]

That's an idea thanks! I'll look for some here and in bigger bore. You think that the beauty of using plastic is not just that it's faster but that it's bendable...

 

It's a shame I paid someone a lot of money and am now having to waste your and my time, plus difficult discussions with said installer. 

 

[MikeSharp01]

2 hours ago, jamesdiyer said:

It's a shame I paid someone a lot of money and am now having to waste your and my time, plus difficult discussions with said installer. 

Its not wasting our time, many of us want to understand this better, and its adding to the community of knowledge that pretty soon everybody using the LLMs will have access to as well. I Just asked Chat GPT and Copilot to run the calcs on this query: 

 

"I have a flow rate of 1000l/Hr in my heating system and 28mm copper pipe work (external Diameter) I need to realign the pipe work using either 2 90degree bends or a single flexible pipe. Can you give me an analysis of both options in terms of pressure drop please."

 

From Co-Pilot I got:

Comparison Summary

 

 

🧩 Design Implications

• The flexible bend offers ~80% lower pressure drop, which is beneficial for pump efficiency and flow stability.
• It also reduces turbulence, which can help with noise and long-term wear.
• If space and installation constraints allow, the flexible pipe is clearly the hydraulically superior option.

 

From Chat GPT I got:

 

Calculated pressure losses — summary

All values shown in Pascals (Pa); 1 Pa = 0.01 mbar.

1. Two 90° long-radius elbows + 0.5 m copper

   • elbow minor losses: 2K(ρv2/2)=2×0.3×(ρv2/2)2K(\rho v^2/2)=2×0.3×(ρv^2/2)2K(ρv2/2)=2×0.3×(ρv2/2) ≈ 80.7 Pa

   • friction (0.5 m): ≈ 61.3 Pa

   • Total ≈ 142.0 Pa (≈ 1.42 mbar)

2. Two 90° short-radius elbows + 0.5 m copper

   • elbow losses ≈ 242.2 Pa

   • friction (0.5 m) ≈ 61.3 Pa

   • Total ≈ 303.5 Pa (≈ 3.04 mbar)

3. Two very sharp 90° bends + 0.5 m copper

   • elbow losses ≈ 403.6 Pa

   • friction (0.5 m) ≈ 61.3 Pa

   • Total ≈ 464.9 Pa (≈ 4.65 mbar)

4. Single flexible connector, 0.5 m long (same ID = 26 mm)

   • friction (0.5 m) ≈ 63.2 Pa

   • (plus small end fittings—typically tens of Pa if any)

   • Total ≈ 63.2 Pa (≈ 0.63 mbar)

Recommendations

• If your flex connector has the same internal bore as the copper and is reasonably short (≈0.5 m), go flexible for slightly lower hydraulic loss and easier alignment.

• If the flex has smaller ID or will be long, use long-radius elbows or re-route with larger-bore flex — avoid small-bore flex for flow-critical runs.

• If you want exact numbers for your exact components, tell me:

  • actual pipe wall thickness or measured ID of your 28 mm tube,

  • the internal diameter and length of the flex you’re considering, and

  • whether your elbows are long-radius or short/standard.

 

So my thoughts are:

 

They are both within a few pascals. They do make assumptions and of course they know nothing about how tortious the flexi will be bent but it looks like, assuming they have both not hallucinated, the flexi option comes with a lower loss BUT if you look at the difference in head loss its only around 25mm (0.025m) between them so not that significant in your situation.

 

[dpmiller]

those fittings are nowhere near 28mm...

[jamesdiyer]

Thanks but yes the internal plastic pipe fittings are a lot less than 28 cooper. 26 id plastic and maybe 20 id fitting. 

 

I'm not going to push for chabing those elbows. What I want is the whole thing done in 35 copper or 40 plastic with compression fittings - don't alter bore. 

 

I cleaned the strainer and went from 1650 to 1750. 

 

I'm sure that going to 32mm id pipe will push is well over 2000l/h. I'm going to work out how to prove it later. 

 

I'm really unhappy with the installers lack of care and though and more so with Vaillant Spain who are useless. 

[jamesdiyer]

.

[Nick Laslett]

6 hours ago, MikeSharp01 said:

In theory yes, but might a couple of large diameter flexis not make all the transitions smoothly and so remove the need for the 90 deg angles everywhere.

 

Just a note for anyone looking at the BES flexible hoses for ASHP. Part Number: 25287, Manufacturer's Part Number: FH-500-28F

 

https://www.bes.co.uk/flexible-hose-1-bsp-f-swivel-x-28mm-x-500mm-pair-25287/

 

For some reason they list them as having a 28mm bore, not sure why? This is just the outer bore, so how useful is that?

 

On the manufacture’s spec sheet they are DN25, e.g. 25mm inner diameter. 

 

https://diversitech.global/storage/app/media/DataSheets/Technical Data Sheets/HOSES/H-FLEXIBLE HOSE-TDS.pdf

 

They also have a DN32 flexible hose, this seems harder to find. 
 

Not really that important, but when I was getting the flexible hoses for my ASHP install, it was frustrating that the inner diameter information wasn’t easier to find. 

Edited Sunday at 14:43 by Nick Laslett

[MikeSharp01]

I did the calcs with 28mm because ours is 28mm so was interested to know the figures and given the size and relative difference I figured it will scale OK.

[jamesdiyer]

51 minutes ago, MikeSharp01 said:

I did the calcs with 28mm because ours is 28mm so was interested to know the figures and given the size and relative difference I figured it will scale OK.

 

Do be cautious using AI for calculations, I've used it for some and it's often missing things and making wrong assumptions. I think your calc is simplistic but when it get a bit more complicated the simple ones fall down. It can be used to check calculations you've done a bit better if you know what to prompt / assume. 

 

 

 

[jamesdiyer]

Do people think that DN25 strainer is a flow block? The mesh seems very wide to be honest. I really can't see that being my issue, but the piping. I'm determined to hit over 2000l/h as I know it's possible. 

 

1750 will be ok. I'm not happy with the installer being a cowboy and the money that was paid. It would have cost not much more to do it in the 40mm it should have been done in to start with, but he doesn't do calculations and that's how he always does it. 

 

Vaillant have to come out for a guarantee inspection/comission so we will see if they know more than nothing. 

[MikeSharp01]

17 minutes ago, jamesdiyer said:

It can be used to check calculations you've done a bit better if you know what to prompt / assume. 

Yes, that's why I use two and if they don't agree I dig deeper. On our system, which I will post for a sense check in a few days, I did it by hand then with 3 LLMs which didn't agree with me so I got them to explain why we differed and then reworked it until we all, well 2 LLMs and I, got to the same point. Which was better than going for my first calc because, it turns out, I had the wrong chart for pressure drop on the pex-al-pex and the TOG rating of the wooden floor wrong. Chat GPT worked best in my view. 

[jamesdiyer]

So, looking at the pump chart a flow of 1750l/hr gives a pressure of ~65kPa. My pressure looks calculations say the 26mm internal gives a pressure of 11kPa, but 32mm gives kPa of 4.5. Thus change to 32mm would reduce presse by ~6.5kPa and bring system pressure to 59kPa give or take. Giving a flow of... 2000l/h. 

 

 

Also the pressure per m of the 26 is ~550Pa/m well over the guidelines of 350Pa/m. A 32mm id would be 219Pa/m

 

[JohnMo]

1 hour ago, jamesdiyer said:

Do people think that DN25 strainer is a flow block?

Trouble with them is the available flow area is pretty low, especially with any debris collected. As I said earlier I had a good jump in flow removing the element and installing a proper filter.

 

But the improvement in flow is really dependent on where your system has the bottle necks. You got a 5% increase in flow, just by cleaning and that is the issue, any debris in the system the system unseen just slows down. 

 

1 minute ago, jamesdiyer said:

So, looking at the pump chart a flow of 1750l/s gives a pressure of ~65kPa. My pressure looks calculations say the 26mm internal gives a pressure of 11kPa, but 32mm gives kPa of 4.5. Thus change to 32mm would reduce presse by ~6.5kPa and bring system pressure to 59kPa give or take. Giving a flow of... 2000l/h. 

 

 

But as said above you need to identify where you actually have bottle necks. 26 to 32mm may save you pressure drop but is the issue else where, if so it will not help you. You need to go through the whole index loop and see what is what. Not just pick an area and point at that, as the issue.

[jamesdiyer]

12 minutes ago, JohnMo said:

Trouble with them is the available flow area is pretty low, especially with any debris collected. As I said earlier I had a good jump in flow removing the element and installing a proper filter.

 

But the improvement in flow is really dependent on where your system has the bottle necks. You got a 5% increase in flow, just by cleaning and that is the issue, any debris in the system the system unseen just slows down. 

 

But as said above you need to identify where you actually have bottle necks. 26 to 32mm may save you pressure drop but is the issue else where, if so it will not help you. You need to go through the whole index loop and see what is what. Not just pick an area and point at that, as the issue.

 

Okay thanks. What filter did you install? 

 

 

I see your point on the index, but the system is quite simple, as sketched previously. Immediately splits into two 22mm coppers after the 3way diverter, so flow halves, then it runs 22mm with 15mm tees to 6 radiators on each of the 22. So I cant see how having what I felt was the index circuit I drew earlier being the issue; that index circuit can't really change, but the new installed pipe could have (should have) been different - to try get max flow. So I'm pointing at the primary as it seems the issue to me. Plus the guidelines say it's an issue as pressure per m is too high and to achieve 15kw at DT 5 isn't possible. I'm happy to be wrong. But to me everything screams that the install of the new pipe was not done to best practice - and thus is wrong. If flow is king a 15kw machine demands bigger bore pipe. 

Edited Sunday at 18:59 by jamesdiyer

[Dillsue]

1 hour ago, jamesdiyer said:

Do people think that DN25 strainer is a flow block? The mesh seems very wide to be honest. 

At a guess I'd say our IntaKlean filter mesh is 2-4 times the surface area of the mesh in your picture, so my view would be that your strainer is potentially too small. Have you checked if there's one fitted internally in the HP?

 

The Vaillant hydraulic schematic I've just looked at has a mag filter in the return leg from the CH

[MikeSharp01]

39 minutes ago, jamesdiyer said:

1750l/s

That feels too big, do you mean l/hr?

[jamesdiyer]

3 minutes ago, MikeSharp01 said:

That feels too big, do you mean l/hr?

Yes of course sorry. Can't edit it now. 

Edited Sunday at 19:24 by jamesdiyer

[MikeSharp01]

6 minutes ago, jamesdiyer said:

9 minutes ago, MikeSharp01 said:

do you mean l/hr?

Yes of course sorry. Can't edit it now. 

That's all done.

[JohnMo]

43 minutes ago, jamesdiyer said:

What filter did you install?

Actually installed a spirodirt filter

[jamesdiyer]

7 minutes ago, JohnMo said:

Actually installed a spirodirt filter

Thanks, I'll email them. That's better as Dutch and available here. The brands mentioned earlier and British and don't distribute in Europe. 

[sharpener]

On 17/10/2025 at 18:11, jamesdiyer said:

On another note, my monitor says basic circuit diagram 8. However if I am just with a volumiser and DHW tank, radiators only 1 circuit, should this be 7? 

 

No it should be 8. This setting is nothing to do with the numbering of the options in the schematics file.

 

8 is correct for all systems without buffer vessels (not including volumisers). Buffers need setting 10. Other numbers are for boilers and hybrid systems.

 

Re pressure drop, your system is sufficiently close to mine (which has 28mm o/d copper primaries so about 26mm i/d - and a lot [?12] of right angle bends) that I think this is not the problem and would look to the filters. IMO you should have a decent branded large area magnetic/mesh filter and NOTHING ELSE. 

 

I have a Y-strainer like yours in my rainwater harvesting system and a hard-to-see tenacious film forms across the gauze surface which causes quite a pressure drop.

 

 

Edited Sunday at 21:24 by sharpener

[SimonD]

3 hours ago, jamesdiyer said:

So I'm pointing at the primary as it seems the issue to me

 

But you've said you have 2000l/h running dhw. If the primary was the problem, you'd see dhw problems too. So to me it's an indication that you have to investigate the whole index circuit, as suggested earlier. 

 

 

[jamesdiyer]

6 hours ago, SimonD said:

 

But you've said you have 2000l/h running dhw. If the primary was the problem, you'd see dhw problems too. So to me it's an indication that you have to investigate the whole index circuit, as suggested earlier. 

 

 

 Unless the primary can run a lot higher flow than 2000? Which I would expect to see (unless limited to max of 2050, so see DHW on theoretical >2050 and heating on 2050. The heating circuit is always going to be lower flow than DHW due to the increased pressure drop, so the system need be designed so it can flow in that and hit a required flow rate, and the DHW flow is just secondary and whatever it is... No? 

 

But in your scenario we can come back to the 25mm pipe that was installed between the 3 way and heating and heating and volumizer on return leg. Which will be changed when they're back. 

Edited yesterday at 05:00 by jamesdiyer

[jamesdiyer]

Vaillant Spain technical head called me. Confirmed 1750l/h isn't enough or acceptable. Said the tube is too small. They have to come out and do an inspection/commissioning for guarantee, so when he comes they will look and then speak with installer.