Anyone else with a monitored heat pump and a Mixergy cylinder?

Zarch · 2 July 2023 14:22

I was wondering whether anyone else with a heat meter monitored heat pump has a Mixergy cylinder attached?

I’ve got a 250L direct version with both the heat pump plate kit and also the PV divert module so I can run the myenergi Eddi against it. So no internal coil, just an external plate heat exchange.

It’s running alongside my Vaillant Arotherm 5kW heat pump.

https://emoncms.org/energystatsuk

Reason I ask if anyone else has one is that there’s a little more to tinker with control wise.
ie, there’s the added complication of the pump speed on the clyinder side, as well as the whole Mixergy % full unique selling point.

Plus, you have to trick the heat pump controller into knowing about the Mixergy on the heat pump controls (in conjunction with the Mixergy heat pump little control box)

My cylinder has a Lowara ecocirc M 15-6/130 N pump. Not sure if everyone has this pump?

Back and forth with Mixergy and their standard answer is “stick it on fixed speed 3”.

I’ve also been tinkering with the pump speed on the Vaillant side.
Should the flow rate on vaillant side match the cylinder side?

No one, including Mixergy or Lowara can tell me the flow rate at 1,2 or 3 pump speeds.
I wish i’d had a flow rate meter put in on that side.

I’ll be honest, even after a full winter i’m not sure if i’m getting the best COP out of it.

Here is this mornings run (this is Auto flow rate on the Vaillant side, pump speed 3 on the Mixergy)

Which was around 45% to 100% full at target 50C.

I need to do more investigation, but anecdotally it does seem its better at long hot water runs than short top ups… but I think that is true for most heat pumps?

if you have a Mixergy, please let us know how you’re getting on with it.
Or whether you’ve got any tips/tricks how to get the best out of it alongside a heat pump.

TrystanLea · 4 July 2023 11:33

Thanks for posting this @Zarch, interesting to see. COP of 3.4 getting the cylinder to 50C seems quite good. I got 3.97 getting to 42C yesterday with similar outside temperature’s to you, not a mixergy cylinder though:

Long runs to get a longer period of time at those lower end temperatures seems to give better results from my experience as well…

Experimenting with different secondary flow rates with the heat exchanger simulator that I put together a while back Heat pump & heat exchanger calculator suggests that there’s not a huge difference with the secondary flow rates… a slight improvement as you increase the flow rate but that doesnt take into account the increased pump power…

Zarch · 4 July 2023 11:49

No worries Trystan. I seems like most people are focused on maximising heating COP, but sort of forget the DHW side. And really, I got lucky on that run as it’s probably the best COP i’ve seen on DHW. I’ve barely done any ‘hot’ day DHW cos i’ve just been ploughing free sunshine into the cylinder!

For the moment, 50C is the lowest target temperature that Mixergy allow.
I have raised a ticket with them to see if they will consider dropping this to 45C (just so I can do some testing)

They seem to want to be in control. They recently dropped the max temp from 65C to 60C in the app citing customer feedback that the water was too hot!!

This was mainly as you can crank the immersion dial and get really hot water via PV divert (hotter than you can set the heat pump to get to in the app).

Bit of a nanny state if i’m being honest.

Personally, Damon put my a solar thermal type mixer in front of my cylinder, so all outlet water is around 46C. So I take my tank upto the 65C max via the immersion/PV divert whenever I can.

Zarch · 5 July 2023 18:15

This has been a painstaking effort in data collection as we don’t currently have a way of pulling out hot water only data from Openenergymon yet.

I went through every day from November through to end of April 2023 collecting just the hot water data.

Electric in, heat out and daily COP of those runs.

Here is the result.

Overall DHW COP across the whole 6 month period - 2.7

What shocked me is there is little consistency in the COP as it can be 2.5 to over 3 for same outside temp on different days.

Granted, I have been playing a lot with settings and had some issues with my original heat meter that restricted flow rates that I could use.

So some of the runs could have been anywhere between target 50C and 65C.
Flow rates from 500 upto 900.
Some could have been small top ups, some runs almost full cylinders.

I have no idea how this compares to other heatpumps, other Arotherms or other Mixergy cylinders.

Fingers crossed we can come up with a DHW only solution for pulling out just that data.

Timbones · 6 July 2023 10:57

I’ve been recording energy for heating and hot water separately from my Mitsubishi Ecodan, so I can do a similar analysis over the same period. (Cylinder is not Mixergy, but does have external PHEX).

COP for hot water of 3.15 for those 6 months, average outdoor temp of 7.1°C, heating tank to 43.7°C.

I’ve seen that COP for hot water cycles is hugely dependant on how hot the water is in the tank when it runs. I feel I get the best performance by only running DHW cycle once per day around midday.

TrystanLea · 20 July 2023 18:58

Where is the inlet and outlet positioned on the tank for both of your HEX based systems?

Your results look similar Tim to Mick’s with a slight performance improvement due to lower target tank temperatures?

Is one of the issues with HEX based systems that the water ends up mixing more in the tank, disturbing the thermocline? what should the result of that look like in the data I wonder?

Interesting comparing the shape of the heat up period up to 52.9C on one of your recent cylinder heat up’s Mick:

Yours (part run clipped to finish at 52.9C flow)

Mine (complete run, finishing at 52.9C flow)

I wonder why the max heat output peaks earlier on on yours and then seems to decay a bit?

TrystanLea · 20 July 2023 19:03

Another comparison, slightly higher outside temps, flow rate shown:

Micks:

Mine:

Timbones · 20 July 2023 22:55

Grey rectangle with 4 pipes is the HEX, in the bottom half of this 250L cylinder. Hot flow from HP enters top left, return is bottom left. I therefore assume that top right takes cold from the bottom of the tank, and bottom right is the hot water that goes up through shiny descaler and pump, and re-enters the cylinder about half-way up. Full picture here.

I can only guess what the flow pattern is like within the cylinder: if the returning water isn’t as hot as what’s already in the middle, then it will sink to the bottom and go round the HEX again. When it gets warmer than what’s in there, it will convect upwards, pushing the cooler water downwards?

I’ve just noticed that the circulation pump between the cylinder and the HEX has 3 speed settings, and is currently set to “2”. Is this worth messing with, or should I leave it well alone?

Here’s a recent cycle (today) from mine to add to the comparisons, up to 53C.
Temperature of the tank started at 18.5C, and reached 46C at the end of the chart.
dT was between 4 and 5.

I also have a second thermister pocket a little higher up in the tank that’s not being used. Wonder if it’s worth poking something in there to collect more data?

Timbones · 21 July 2023 09:31

To answer my own question, the manufacturer’s installation instructions explicitly says:

Ensure the DHW circulation pump is set to speed setting II to ensure optimum DHW heating performance. Failure to do so can result in excessive use of the back-up immersion heater or lower storage temperatures than required.

Zarch · 21 July 2023 13:43

Purple = flow/return on the heat pump side into and out of the plate.
The red circles indicate entry/exit points on the cylinder.

You can see here from Mixergy manual

Zarch · 21 July 2023 13:47

This for me is a massive point and one i’m trying to make to Mixergy.

Here is a copy/paste of what i’ve sent them this morning.

Max flow rate on the heat pump side for me with my 5Kw Arotherm is around 860lph and as we’ve had conversations before, we don’t know what the flow rate of the plate pump is on any of the 1,2 or 3 settings.

As heat geek explain on this video, unless you get the flow rate the same on both side of the separation (low loss headers or plate heat exchanges) you will get distortion which leads to lower performance.

https://www.youtube.com/watch?v=mNcRx45DQ8M&ab_channel=HeatGeek

In the case of my 5kW Arotherm at 100% pump speed, my flow rate is around 860 lph

On the 7kW Arotherm 100% is 1200 lph and on the 10 & 12kW models its 2000lph.

So there is no way that the cylinder side pump can match all three of those flow rates by ‘just put it in setting 3’ can it?

The one stop shop of a single setting on the Lowra pump can’t cover all heat pump types, models and sizes?

To address this, how can we work out or know the flow rate of the 3 settings on the cylinder side?

If we knew what the Lowra flow rates were we could have a better stab at matching the heat pump side to optimise performance.

Was there ever any testing done on the flow rates on the cylinder side pump?

Could we add a flow rate meter on that side to get an idea what they are?

So I think you’re in the same boat as me there @Timbones how do you what the flow rate is on the cylinder side of the plate to match it up to the heat pump side flow rate (which you know or could fix) if you knew the cylinder side?

TrystanLea · 21 July 2023 19:06

I wonder how important that matching really is?

Using the Effectiveness NTU method to simulate a heat exchanger Heat pump & heat exchanger calculator, in that particular case connected to radiators:

if we set the primary flow rate to 12 L/min, best performance is of course with matched flow rate, 12 L/min on secondary COP with parameters in the default example 2.82
If I reduce the flow rate on the secondary to 6 L/min the COP only drops to 2.80
If I increase the flow rate on the secondary to 24 L/min the COP only drops to 2.81

Those changes are in the margin of error of the heat meter. Perhaps the calculation behind that isn’t really capturing the reality of what happens in the heat exchanger?

I wonder if it’s the combination of the distortion caused by mixing in the tank and the distortion in the heat exchanger that reduces performance?

Zarch · 22 July 2023 09:16

This is all certainly going beyond my current knowledge.

And from the back and forth with Mixergy themselves, it seems beyond them too at this point.
I’m hoping someone more technical starts to input from their support side.

Bottom line for me, they can’t suggest their plate exchanger is 3msq and it not perform the same as 3msq coils?

Especially when i’ve provided heat meter evidence that other Arotherms (and other heat pumps) with coil cylinders are consistently outperforming my setup.

TrystanLea · 22 July 2023 10:03

Yes conclusion seems fair!

@johncantor and @Timbones both shared some interesting DHW heat up graphs on the twitter discussion.

First John’s graph, notice that very close return temp! Flow temp is 7.5 degrees above cylinder temperature?

Tim’s graph, tank temperature ~3-4 degrees below return temperature:

Here’s mine for comparison
Not actually quite as good as John’s example above on the return:

First a low power run ~ 2290W average heat

Flow temperature is typically 4.5K above cylinder temperature, Return temperature about 1.5K above cylinder temperature.

Then at higher power ~4950W average heat

Flow - Cyl = 8K
Return - Cyl = 1.5K

The difference in Return - Cylinder temperature seems to be fairly consistent at 1.5K on my system when I check other runs. The flow return DT being proportional to power as I have a fixed speed pump…

Zarch · 22 July 2023 10:06

Funnily enough, i’m just setting up home assistant triggers to push Mixergy hot and cold temps into my emoncms so I can then graph it all in there.

As well as flow/return from the heat meter I have put separate temp probes on bits of the HEX (both sides).

TrystanLea · 22 July 2023 10:07

Great, look forward to seeing that!

Zarch · 22 July 2023 10:35

I’ve obviously got “real” flow and return logging into emoncms from the heat meter.

But because there is always a slight temperature variation versus ds18b20 attached to pipes I have also attached as many of these ds18b20 to various points as I could. Just taped to the pipe under the insulation with some thermal paste attached.

Flow
Return
ASHP side of plate (Flow)
Tank side (into tank)
ASHP side of plate (Return)

Note: there isn’t really anywhere to plonk a sensor on the (out of tank) pipework as it’s so short. Plus, i’m all out of emon connections for more ds18b20.

Although I wondering the usefulness of the ‘into the plate’ as it’s only a couple of feet from the flow ds18b20

Also getting the following info.

Mixergy capacity (% - from the tank itself)
Mixergy Hot Water Temp (from the tank itself)
Mixergy Cold Water Temp (from the tank itself)

Note: the above three are pulled into Home Assistant from the very excellent Mixergy integration by Tom McGuiness. I then forward those readings at 1 min intervals onto emoncms so I can graph it all in one place.

Zarch · 23 July 2023 09:40

Here’s yesterday afternoon hot water run.

It was starting from 30% full to 100% (of a 250L cylinder) with target temp of 50C.

Bizarrely, I think this was the best COP i’ve ever seen on hot water, 3.53.

And here is the sensor graph with all the new data over in emomcms. Wow, lots of data.
@TrystanLea this is all on my zarch emoncms account.
Would it be better if this was all over on my other public account?

So, how do we make sense of that and can we think of any changes off the back of it?

Timbones · 24 July 2023 09:35

I question how “setting II” can be right for all possible installations? I do know that the (fixed) flow on the heat pump side (15 l/min) is at the very bottom end of the manufacturer’s spec sheet (14 - 32), so it seems reasonable to think that a lower setting would be a better match for my system. However, there’s no way to know without measuring the flow of the cylinder side, or doing detailed experiments. Perhaps this is why instructions say to fix it at a particular setting?

Very interesting, thanks. This suggests that the pump speed doesn’t matter, though why does manufacturer then warn that an incorrect setting would result in worse performance?

Would it be possible to extrapolate flow rates through the HEX from the spec sheet of the pump, assuming they’re trying to match the recommended flow rates of the heat pump?

Zarch · 24 July 2023 11:03

I’ve asked the very same question to Mixergy about the Lowara pump they use. They can’t answer and there is nothing in the manual around specific flow rates. I even emailed Lowara directly and received no reply.

I think the only way would be to put some sort of flow meter / flow rate read-out dial inline after the pump and before the cylinder?

Adam from heatgeek keeps telling me the problem is ‘distortion’, but Trystan’s hex calculator suggests it shouldn’t be that big a deal, so not sure what to believe?

But i’m with you and make this very same point to Mixergy, especially as they use the same pump and same plate for all heat pumps, you don’t have different ones for different output units.