I think the difference between testbed and realworld here vs others is that in say cars, you could never get the MPG advertised (although now there are more realistic test standards that give usable (ish) numbers).

But in this case the HP testbeds are all openloop (its not a secret) and you CAN get those numbers if you build a real world system that way.  if you build with a buffer or llh or phe you will get a few % off parity at best, if it is engineered well (or you get lucky). If not, then a lot of % off. clearly we have both ends of that spectrum here, some are happy (ish), some are unhappy. 

I also studied lots of vendor schematics and documents, and publications and youtubes from people who know what they are talking about . none of them say "you must have a buffer/llh/phe". the only reason the situation that has come about with them still being put in by default, is, as far as I can tell, is as a legacy of the earlier installer "rules of thumb" where they were all paranoid about getting flow rate issues causing the HP to error, so they'd mitigate that with one of these devices, and somehow its stuck. Only certain installers (plus most/all of the DIY'ers) are confident enough to go against this.

Take a look at the heatpumpmonitor.org system list, set "hydraulic separation" to on in the display, and sort by COP. The distribution isn't perfect, and there are plenty of other reasons why one system can perform better than another as we know,  but there is still a significant bias towards those with LLH/Buffer being lower down the list and those with None being higher. QED.

the other issues mentioned as reasons to keep these devices , aren't. dirt - use a suitable low resistance filter. freezing - antifreeze valves.

the only reasons IMO to keep / put in those devices in a domestic setup are if

• its already in, your system is performing reasonably well and the cost/hassle of removing it for a slightly better performance isn't worth it
• you have more complex system than usual which genuinely requires hydraulic separation . e.g. multiple heat sources. A real "buffer tank" (i.e. a big one , capable of storing a lot of heat) has a use case here as well, if one heat source is uncontrolled. 

Spot on IMHO and my studying, like yours, of manuals, posts and the various other info available.

Posted by: @jamespa

↑

Thats from Freedom, what does the OEM actually say

Here is the system diagram from page 4 of the Midea manual.  No buffer tank, LLH or PHE in sight.

This is a variation on efficacy vs effectiveness. That's what Midea say in their 'ideal' manual. In the real world, stuff happens, Freedom offer training courses and provide installation manuals, and that is what the installers use. Why wouldn't they? Freedom are a major supplier, surely they know what they are talking about?

You will have to ask Freedom why they varied the spec from the Midea manual.

Posted by: @jamespa

↑

Spot on IMHO and my studying, like yours, of manuals, posts and the various other info available.

But it is just unrealistic to expect most people in the real world to do this. And there is a related question, should they be expected to do it?

Posted by: @cathoderay

↑

But it is just unrealistic to expect most people in the real world to do this. And there is a related question, should they be expected to do it?

no of course end users shouldn't be doing this. Installers should. Installers should be MADE to put in systems that perform properly and held liable (not paid) if they don't (perform properly). 

In your case, freedom put their own rules around the system, in to protect themselves from "low flow" callbacks. plus someone underspecced the PHE. So you've ended up with a system design that would have worked within a few % if it was well engineered, but its NOT well engineered (the aforementioned PHE). Bizarrely you might have actually done better with an LLH! This is all unfortunate obviously for you and you have my sympathy, but doesn't change my view on the bigger picture.

Posted by: @iancalderbank

↑

Take a look at the heatpumpmonitor.org system list, set "hydraulic separation" to on in the display, and sort by COP. The distribution isn't perfect, and there are plenty of other reasons why one system can perform better than another as we know,  but there is still a significant bias towards those with LLH/Buffer being lower down the list and those with None being higher. QED.

Interesting, I've just done what you suggest, but I don't think it gets near a QED threshold, and as we all know, and you say ('plenty of other reason'),  association isn't causation.

I can absolutely see how a PHE (which is what I have) throttles output, it can only transfer what it can transfer, and is the throughout is less than that needed, then the system is throttled. But I am not sure that it necessarily impairs efficiency.  

Posted by: @iancalderbank

↑

plus someone underspecced the PHE

I'm pretty sure this was Freedom. In effect, as I understand it, they supply an installation kit, and the installer installs what is in the kit. I've mentioned this before, I am pretty sure I asked my installer at installation time how the PHE was sized, and I think - it's two years ago now - that his reply was Freedom supply a generic one, large enough to cover most installations. 

I do wonder if part of the problem is that heat installation as done in this country seems to be largely done as a trade, and trades persons don't as a rule get involved with serious (engineering) design, only basic stuff, which heat pumps definitely are not. This creates a vacuum into which suppliers can step: here's our half day training course, we'll have you up to speed in no time, we'll even give you a kit of parts and an installation manual, and Bob's your uncle. Add in a gutless toothless regulation system, and before you know it, we have what we have.

Again, it is back to the real world. Most car mechanics don't have degrees in physics, and neither do most heat pump installers (so do, I know). It would take a huge effort to change that, and off the top of my head, right now I can't see a way to do it. 

Posted by: @cathoderay

↑

 But I am not sure that it necessarily impairs efficiency.  

It does. I do not claim to be the world authority on PHE's but if you read the public explanations of how they work, full efficiency running a HP across a PHE is physically impossible in the context of what we're talking about - you can't get the output hot to exactly the same temperature as the input hot. you can move all the heat energy, but it'll always be degraded a bit on the output side. which must have cost you some efficiency in the HP.

whereas a straight piece of pipe doesn't have any system loss just a tiny bit of thermal (but we're assuming both systems are well insulated so that negates).

It gets harder to engineer, the smaller the DT, the bigger PHE you need. The issue with yours is more that its undersized. if it had been sized right we'd probably not be having this debate, the few % inefficiency would be a non-issue because your house would be warm.

Posted by: @cathoderay

↑

I do wonder if part of the problem is that heat installation as done in this country seems to be largely done as a trade, and trades persons don't as a rule get involved with serious (engineering) design, only basic stuff, which heat pumps definitely are not. This creates a vacuum into which suppliers can step: here's our half day training course, we'll have you up to speed in no time, we'll even give you a kit of parts and an installation manual, and Bob's your uncle. Add in a gutless toothless regulation system, and before you know it, we have what we have.

Agree.  Doubly so with retrofits, where system design in the context of the specific house is the key to an implementation which is effective, cost effective, and efficient.  Of course there are good suppliers as well as good installers, but sadly not sufficient at present.

Posted by: @cathoderay

↑

Again, it is back to the real world. Most car mechanics don't have degrees in physics, and neither do most heat pump installers (so do, I know). It would take a huge effort to change that, and off the top of my head, right now I can't see a way to do it. 

I don't think you need a degree in physics (although it probably helps), but you do need engineering skills to do the system design and problem solving.  Many plumbers are essentially fitters, very skilled at what they do but not able to do the engineering.  Unfortunately the system engineering and problem solving associated with plumbing, which I think is inherently interesting, comes with the requirement to bend yourself double in a confined space, whilst being looked down upon by a sector of the British public.  Not the greatest incentive for a bright young graduate.  

Posted by: @cathoderay

↑

Freedom are a major supplier, surely they know what they are talking about?

An interesting assumption, to say the least!  But lets suppose its true, you still have to question what they optimise for.  

Posted by: @iancalderbank

↑

Posted by: @cathoderay

↑

 But I am not sure that it necessarily impairs efficiency.  

 

 

It does. I do not claim to be the world authority on PHE's but if you read the public explanations of how they work, full efficiency running a HP across a PHE is physically impossible in the context of what we're talking about - you can't get the output hot to exactly the same temperature as the input hot. you can move all the heat energy, but it'll always be degraded a bit on the output side. which must have cost you some efficiency in the HP.

whereas a straight piece of pipe doesn't have any system loss just a tiny bit of thermal (but we're assuming both systems are well insulated so that negates).

It gets harder to engineer, the smaller the DT, the bigger PHE you need. The issue with yours is more that its undersized. if it had been sized right we'd probably not be having this debate, the few % inefficiency would be a non-issue because your house would be warm.

If there is a temperature drop flow-flow across the PHE/buffer/LLH it must impair system efficiency because it requires you to increase the flow temperature from the heat pump to achieve the necessary flow temp at the emitters.  System efficiency drops with increased flow temperature, its in the thermodynamics and the manufacturers data tables.  

Posted by: @iancalderbank

↑

It does

I'm still sure about this, I think we might be mixing up throughput with efficiency. Consider the PHE as a form of obstruction: it restricts throughput. Even if the heat pump can produce 11kW, and the PHE can only throughput 8kW, then you have, in effect, my situation, with the heat pump reacting by only producing 8kW. A bit like trying to blow down a very small bore pipe. Your lungs can produce enough airflow to blow out candles six feet away, but the small bore pipe restricts the flow, and you only can only blow out one candle three feet away. But does that mean your lungs are now less efficient? Less capable, certainly, but less efficient? I can imagine something similar with a heat pump: its output is restricted, but it can still produce that lower output efficiently. 

Posted by: @jamespa

↑

comes with the requirement to bend yourself double in a confined space, whilst being looked down upon by a sector of the British public.  Not the greatest incentive for a bright young graduate.  

Nicely put.

Posted by: @cathoderay

↑

Yes, I get that, but how large is that loss percentage-wise? The throttling effect reduces say 11kW to 8kW - a 27% reduction. I suspect the blockage is the main problem, not the agreed but possibly not very large drop in efficiency. Again, I am distinguishing between throughput on the one hand and efficiency on the other.

As stated upthread about 2-3% per deg C, but if you want an accurate number check your systems data tables.

Throttling is a different phenomenon which may (will) cause a further loss in efficiency also because water is being recirculated through the heat pump without doing anything useful and you cant get owt for nowt.  However I dont have a figure for this effect.

Posted by: @jamespa

↑

An interesting assumption, to say the least! 

It was a rhetorical question (? at the end), not an assumption!

Posted by: @jamespa

↑

If there is a temperature drop flow-flow across the PHE/buffer/LLH it must impair system efficiency because it requires you to increase the flow temperature from the heat pump to achieve the necessary flow temp at the emitters.

Yes, I get that, but how large is that loss percentage-wise? The throttling effect reduces say 11kW to 8kW - a 27% reduction. I suspect the blockage is the main problem, not the agreed but possibly not very large drop in efficiency. Again, I am distinguishing between throughput on the one hand and efficiency on the other.