Just an observation regarding the Joule insulated 200ltr unvented hot water tank, is this tank designed with heat pumps in mind? Though probably sufficient capacity as you have been using it until now, the internal pipework may not be of sufficient capacity to heat your water at the lower temperatures that a heat pump provides. I am bearing in mind that you mentioned a 600 x 600 mm standard space and a new tank large enough for your needs may not fit in the existing space. Regards, Toodles.

Thanks for the question I tend to use the electric emersion heater more than using the boiler hot water feed I turn on the emersion heater every night during the cheap rate from Octopus intelligent go tariff and I have the Nest controller turn on heating only for 1 hour late afternoon. As there is only 2 people in the house and we only have showers I have never run out of hot water. my intent is to replace my solar panel system with a new one with more panels bigger inverter and 20kw of battery. so eventually I would only rely on electricall heating the cylinder. I agree with your point and i may consider changing the cylinder as i do have an optional price for a 170 ltr combined unit. my big concern is the heat exchanger circuit with the plate heat exchanger as the min volumn required by the heat pump is 40ltr the plate heat exchanger uses 20 ltr and the pipe run to the heat pump/heat exchanger has about 10 ltr making the total volumn of glycol 30Ltr which to me seem far to low.

Posted by: @paulc

↑

as an electrical engineer

OK

Power (Watts) = Flow (in l/s) * Heat capacity of water (4180 J/°C/L) * DeltaT (in °C)

Flow (in l/s) = Power (Watts) /( DeltaT (in °C) * Heat capacity of water (4180 J/°C/L) )

Therefore, 10kW gives 0.5 l/s @ DeltaT 5°C or 1.72 m3/h

Pressure drop in 20mm ID pipe, per meter of pipe

120.0 l/h 0.00 mCE/m
306.7 l/h 0.01 mCE/m
493.3 l/h 0.02 mCE/m
680.0 l/h 0.03 mCE/m
866.7 l/h 0.04 mCE/m
1053.3 l/h 0.06 mCE/m
1240.0 l/h 0.08 mCE/m
1426.7 l/h 0.11 mCE/m
1613.3 l/h 0.13 mCE/m
1800.0 l/h 0.16 mCE/m

Pressure drop budget from your circulator ~4 mCE = 0.4 bar

In other words 22mm OD copper (20mm ID) is too small for 10kW at deltaT5 but it would be okay at deltaT 10°C.

Note pressure drop is in velocity² so flow squared, and pipe ID to a power of about -4.5..-5. Halving pipe ID divides area by 4, so it multiplies velocity by 4, velocity is squared, so half pipe ID multiplies pressure drop by 16, plus extra for turbulence, so maybe 20-30x. That's why when the pipe is too small... it's too small.

At the same power, doubling deltaT halves flow, so it divides pressure drop by 4. Therefore, increasing deltaT is a way to make it work if your pipes are too small. The cost is a small drop in COP since the heat pump has to output a higher temperature.

If you use weather compensation properly, the heat pump will modulate its output flow temperature, always keeping it at the minimum required, for best COP. This means you remove all or most TRVs and only balance rads, and let the heat pump do its thing.

Basically COP drops as flow temperature rises, so anything that causes flow temperature to rise is bad for your bank account. Thus high flow with colder water is cheaper than low flow with hot water. The installer needs to do per room heat loss to check if the rads are sized properly and will work at lower temp. If your rads were sized for 60°C water before you insulated the house, they are now oversized for the insulated house, so they require cooler water, which is exactly what you need.

Posted by: @paulc

↑

The proposal is to install a 10 kw Heat Pump using the R32 Midea mono air-to-water unit.

The only reason to use R32 (which will be outlawed during the lifetime of the machine) is if you get the machine at a significant bargain and DIY the whole installation at minimum cost. Otherwise, R290 (propane).

Posted by: @paulc

↑

In addition, they have proposed to install a Heat Saviour unit

It's a very small copper brazed stainless plate heat exchanger, which you can buy off amazon for less than €100. Are then selling it for more than 2000? The packaging is neat though. Plate heat exchangers (PHX) --

Advantages:

- can use two different fluids, like water with antifreeze on the heat pump side and rusty sludge on the radiator side.

- can have different flow rates on both sides

Drawbacks:

- you will lose a few °C so the heat pump has to output hotter water at lower COP ; the smaller the PHX plate area the worse this is

- from the picture it looks like copper brazed stainless and not full stainless, therefore if your heating fluid is rusty sludge the brazing copper will corrode and develop pinhole leaks after a few years. If your heating fluid is clean, no issue.

I inherited an old installation with ~120m² old UFH loops using non barrier PEX. Therefore oxygen leakage through the non barrier PEX completely destroyed all steel components. Most of the innards of the old boiler is currently inside the UFH pipes in the form of rust flakes. 

At some point a heat pump was added. The issue with this water is the combo of oxygen ingress, tons of dissolved garbage, rust flakes, acidic pH, high conductivity, etc. It perforated the PHX copper brazing in 4 years. Thankfully I had bought the heat pump used for 200€ so didn't care.

Now I have a brand new 16kW A2W heat pump. I don't intend for it to get perforated, so I purchased a second hand stainless/EPDM PHX (previously used for heating an olympic swimming pool, so it is enormous) at the cost of scrap metal and put it between the new installation, which is all copper and MLC pipe, free of oxygen, with heat pump, more brand new UFH and brand new radiant ceilings... and the old cursed non-barrier PEX UFH.

I lose a few °C on the heat exchanger, but I have no choice. I looked for PHX specs in the Heat Saviour website, didn't find any. It would be nice to know the plate area and pressure drop as that are the most important metrics. Without data I would not buy.

What you need to do is take a sample of the water from your installation (say 1 pint).

- Take a picture of the sample inside a white container and post it

- Check pH, conductivity and TDS (using TDS meter and pH strips on amazon)

- Strain through coffee filter

- Post macro picture of particulate matter, if any, with ruler for scale

- Check how much brown (non magnetic) and black (magnetic) particles there are.

Also check strainer or dirt/magnetic filters.

This will sort it out between the three following cases:

- pH>8 (good), clear water with a small amount of particles: your installation is clean, congratulations, no extra cost, just connect the heat pump, use antifreeze drain valves, don't bother with glycol

- pH>8 (good), slightly brownish with a bit of particles: same + power flushing

- Acidic pH, rusty chunky sludge : Desperate, you need a heat exchanger and if it is copper brazed it will be a wear part. Do not connect heat pump directly or its PHX will become the wear part.

More or less needless to say I agree with what @bobflux says, but there are some issues you should sort before you proceed.

Posted by: @paulc

↑

Total heat loss is about 5.7Kw 

confuses me.  If this is the case then a single 22mm OK at DT5 (just) and why are you putting in a 10kW heat pump and messing about with unnecessary PHEs.  Does your installer know what they are doing?

If your loss is in fact 10kW then:

Posted by: @paulc

↑

This is a 3 pipe pressurised system for the heating and hot water using 22mm pipework for the main feeds

Is it a single pair of 22mm pipes one flow and one return, or two pairs (eg a pair for up and a pair for down)?

Are there a few radiators that could be separately plumbed or where you could easily run a parallel feed?  Perhaps there is a way round other than the Heat Saviour or increasing DT with only a small amount of replumbing.

Posted by: @paulc

↑

The Heat Saviour unit

... was a candidate for a RHH Turkey of the year award a couple of years ago, but in the end the award wasn't made

Posted by: @paulc

↑

total volumn of glycol 30Ltr

You should be aiming for a direct connection from heat pump to emitters without an intermediate heat exchanger, buffer tank or LLH unless there is a genuine reason for it (which is rarely the case in a domestic property).  I would also eliminate glycol in favour of anti freeze valves but opinions differ on this.

Posted by: @paulc

↑

Thanks for the question I tend to use the electric emersion heater more than using the boiler hot water feed I turn on the emersion heater every night during the cheap rate from Octopus intelligent go tariff and I have the Nest controller turn on heating only for 1 hour late afternoon

Once you get your heat pump it will be more economical to heat DHW using that if you can.  Here there is a good use for a PHE - to plate load a cylinder that otherwise has an insufficient coil size.  That said with low overnight tarrif the payback time will be long vs just continuing to use the immersion.

You are going to want to ditch the Nest as a controller for the heat pump BTW.  External controls, other than those designed specifically for heat pumps (which Nest is not) compromise both efficiency and effectiveness.

I have an old (and listed) stone built cottage that I heat (both space and DHW) using a Midea R32 heat pump (and a plate heat exchanger). It can be done.

However...

Posted by: @paulc

↑

Total heat loss is about 5.7Kw without taking into account the hot water cylinder consumption but I usually heat my water overnight on the cheap rate.

...

The proposal is to install a 10 kw Heat Pump using the R32 Midea mono air-to-water unit.

The proposed heat pump rating is considerably in excess of the heat loss. This is generally not a good idea because (a) it is unnecessary and (b) the higher rating means its lowest output may be too high (check out the Midea Engineering Data manuals, various links in various posts on the forum).

Getting the very best possible estimate of heat loss before the installation is essential. Past gas use can give a reasonable estimate, but oil (which I also had before the heat pump) much less so because the data is too crude - how much oil did you really use per day? You need the per day data, see previous posts on empirical assessment of heat loss. You will also very likely run your heat pump 24 hours a day (the low and steady principle and open loop ie no TRVs, no external controls, as @bobflux says) which may be very different to how you run your oil central heating (timed fast and furious) and this can (and did for me) increase my energy use, chiefly because the house is warmer for longer. 

Never forget that any spreadsheet based assessment of heat loss is no more or less that whatiffery - what if the U value is this, what if it is that? What if the air changes are more or less? There is some evidence that standard heat loss spreadsheets with pre-set U values over-estimate the values for solid stone walls.  

More generally, my hunch is that the proposed setup is much more complicated than it needs to be. My actual heat loss as determined by energy use by the heat pump is around 10kW. The Midea unit I have is rated at 14kW but that is the output on a sunny spring day. At design temperatures it falls to around 11kW (it's all in the Engineering Data Manuals), so it is just OK, both on paper and in practice. The primary pipework between the heat pump and the plate heat exchanger is 28mm but it is plastic pipe so less, but most of the rad circuits are 22 then 15mm copper pipe. Your pipework may well be OK as it is, given your heat loss may almost half mine. You need to check and recheck the calculations.

Note that it seems to be a quirk of Midea R32 Mono heat pumps that they tend not to vary the flow rate, mine spends most of the time at around 1.4m^3/h, and the rest of time at around 1m^3/h. So far as I know, no one has really worked out how the Midea control logic works...

What I don't think you have yet given enough attention to (or if you have, I don't think you mentioned it) is radiator sizing (output). Very generally, as a rule of thumb, a heat pump system needs much larger radiators to achieve the same heat output, because the flow temperature is so much lower compared to a typical oil system flow temperature. There are standard tables that enable you to convert a radiator's quoted delta t 50 (rad to room delta) output to typical heat pump rad to room delta t. The drop in output is not trivial, eg at a rad to room delta t of 30, the rad's output is only half of what it is at a rad to room delta t of 50.

As a rule, I think it is fair to say that we tend to be suspicious of any solution that claims to be a fix all for all your heat pump problems, or they will somehow soup the system up and make it perform better/more efficiently or whatever. The basics of a heat pump system are extremely simple, and anything fancy added to the system may well be unnecessary, and may even make things worse.

Finally, if you are a geek and want to monitor and even control your Midea system (assuming you get one) in great detail, you are in luck, because Midea provide a two wire read and write modbus connection on the wired controller. Other forum threads and posts detail how to do this. 

I see @jamespa has posted while I was writing this, there is inevitably some common ground.      

Thanks all for your observations and usefull advice, I have Just completed a heat loss calculation (i use to do this back in my consultancy days) it comes out to 5.2kw and I intend to remove most of the TRV,s or open them fully. Having investegated this further the proposal is to use R290 in the midea unit at 8Kw, and have the primary circuit plumbed into a new DHW cylinder (designed for heat pumps), with a plate heat exchanger built into the base of the cylinder. The data I have managed to collect about the old standard steelrad radiators are oversized and in most cases 50% more than is needed. The heating circuit will be one single loop fed prom the heat exchanger and if I find that the heat output is insufficient for the radiators then I can replace the radiators at a later date. This is the curret plan but further tweeks may come later as this process of discovery continues.

Thanks again

Obviously it's entirely up to you whether you heard the advice being given, but what you are saying still isn't quite stacking up to the optimum arrangement.  I will ask a couple of questions which you are under no obligation to answer!

Posted by: @paulc

↑

The data I have managed to collect about the old standard steelrad radiators are oversized and in most cases 50% more than is needed

At what flow temperature?

Posted by: @paulc

↑

The heating circuit will be one single loop fed prom the heat exchanger

Why.  22mm is ok for 6kW at dt5 and you confirm your loss is 5.2kW Why bother with the phe and the inevitable loss of efficiency additional pump and additional complexity in fault diagnosis?

James this is something to be confirmed there was a mention of 55 degrees for the primary would you recomend something different

Posted by: @paulc

↑

James this is something to be confirmed there was a mention of 55 degrees for the primary would you recomend something different

I don't understand.  If they are '50% oversized' that must be calculated at a specific ft.

Furthermore if the ft to the rads is 55 then the ft from the heat pump will need to be higher as you have a phe in the system.  Possibly as high as 60.

I assume you are not claiming the bus grant, or alternatively have checked that this heat pump qualifies if operated at a ft over 55.  Most don't.

Its rare to design for a ft from the heat pump over 55 not least because of the rules associated with the bus grant.  45 is more common.  

Again there are missing links which don't quite stack up to a coherent design.  It may be ok, but the missing information suggests it may well not be.

Posted by: @paulc

↑

there was a mention of 55 degrees for the primary

I agree with @jamespa, this is much higher than normal, though not unheard of. Don't forget the basic principle of 'low and steady' where low specifically means a low (primary) flow temperature.

The radiator correction factor is very simple, I gave a link to such a table of such values. From that table, if a rad has an output of 1000W at a rad room delta t of 50, then at a rad room delta t of 30 (which is on the high side for a heat pump), then output effectively halves to 513W. You need to check the rads on a per room basis, making sure the rad's output at relevant rad to room delta t matches or slightly exceeds each room's heat loss.