Posted by: @agentgeorge

↑

The App says it choses the top temperature when OAT is above 12C, my system is running at 36C today and its 6.8C according to Cosy HP. Logic would suggest its using the higher temp, but its not.

anyone got an idea how WC is being implemented by Cosy6?

I don’t know how it is implemented but the two temps you select are the end points at -10c and +12c.  The WC curve is, as far as people can tell, linear between these two points so you should be able to graph it in Excel (for example) to give you an approximate flow temp at the current OAT.  I actually don’t think it’s anymore complicated than this in the firmware.   If you log in to the Cosy Hub you can see the current OAT and flow temp and see how that correlates to your graph.  Also in the Cosy Hub, you can see your 50c set point.  In theory that should match your design temp and it may have an influence on the linearity of the WC curve: I don’t know, but I can believe that if you have 57c at -10c (set through the app) and 50c at -3.4c it provides a third point on the curve.   I have shifted mine to 50c at -10c (instead of -3.4) and that reflects in the app. At temps outside that range I think it keeps that temp so, for example, if it is set at 36c at 12c, then it is 36c at 13c, 14c, 15c etc… ditto the other end of the curve.

@andrewj found this article by heart geeks

https://www.heatgeek.com/articles/how-to-maximise-your-heat-pump-or-boiler-efficiency

it touches a lot of topics, WC is as you described, a line or curve between 2 OATs where flow temperature is altered.

As I have UFH, I suspect my system doesn't go above 37C. I have no blend valve or secondary pump on the UFH. I have also turned the controller off so all the loops are open circuit. To control the temperature response to the different zones, I have balanced the flow regulators, 0.5l/m to 1.5l/m depending on loop length. This gives even temps downstairs when walking between kitchen, lounge, cloakroom, bedroom 4.

The article also mentions to use TRVs to stop a room overheating once the system has been setup and balanced.

Does this article line up with your experiences?

@editor I can detail the concise version of events, if anything strikes a chord I can elaborate

initially declined to quote, which I deduced by changing my answers to the questions, as property being renovated

duh, thats ideal time to hide the pipes and do an invisible install!

I argued for weeks to get them to come out and do a survey

heat loss survey not accurate as they couldn't put down the insulation level I was going to use as it wasn't fitted!

changed the heat pump from 6kw cosy to daikin 8kw without asking

on day of install they noticed id been sent a used heat pump so that was swapped out at 10pm that night!

on day 2 of install, manager turned up and declared I wasnt eligible for BUS grant as the heat loss was greater than the cosy id insisted on.

I showed them my professional heat loss survey which differ from Octopus one; my guess is they go for next size up “just in case” the smaller one would underperform and they would be liable to change the HP

They got 2 radiators wrong size, undersized one Im still coping with and will swap out myself in the summer, the mainbathroom is not warm enough, only fix is to increase insulation in loft above it with Kingspan 120, again at my expense.

They didnt flush the system out before fitting the delicate HP into the circuit, so it clogged up after 2h of use. It took a week and 5 visits from different technicians before someone actually diagnosed this issue correctly and did a power flush.

HP falls over every day with low flow errors, when the cosy pod decides the downstairs is hot enough and shuts the HP down. This shouldn't be logged as a fault condition

Octopus initially said their system wouldn't work with microbore pipework, but not why. I’ve determined its because the 10mm pipework cant supply the minimum flow rate the cosy6 needs of 12l/m. Also 10mm pipework cant supply can only supply 750w/h, not enough for the larger bedroom.

it took 6 weeks of ignoring my phone calls before they agreed to repipe the rear bedrooms to 15mm to support the underperformance in those rooms.

it was only when they asked for sign off of the system and balance of money paid that they took notice of the email trail with 10 points that needed addressing before I could sign off and pay.

This upset the manager as she was getting demerits every time I called in and every week the final payment wasn't made.

she eventually came round in person and we worked thru the list of 10 issues, i compromised on a couple but held out on 8 and got those resolved.

its fortunate that i was a project manager for main stream automotive manufacturers and could negotiate and not be brow beaten 

@agentgeorge I’m aware of the process - in fact, it’s been mentioned by @jamespa and @toodles in this thread a few times.  The approach requires reducing flow temperatures until target temp is reached and then balancing rads to keep consistency across rooms (or to fine tune temps where required.). If it isn’t possible to get the flow temps low enough then you’re on a hiding to nothing; ditto, if you find you can’t restrict a radiator enough.  This leads to using TRVs not as temperature limiters above nominal temp (to account for, say, solar gain or human gain) but to keep the room temperature down.  This introduces multiple zones into the circuit which just reduces efficiency.  

My experience has been a bit problematic in that the radiators are mostly oversized and the flow temperatures can’t be reduced enough.  My design was for 50c at -3.4c but I am actually running at 50c at -10c.  To get the WC curve to use lower temps at milder temps, I.e. -3.4c and warmer, I need to adjust the warm end and drop the 36c at 12c down.  Unfortunately the Cosy 9 doesn’t like that according to Octopus and they’ve also said it can’t modulate below 35c in any case which seems to be the lower limit without big impacts on efficiency.

I have found that I need to run it thermostatically and so have the thermostat in my downstairs toilet, a 2mx1.5m room with a small window and two (well insulated) external walls.  This, by experiment, is the only room that works as other rooms’ heat loss cause under-/over-heating in the rest of the house.  The set temp for the thermostat in that room at 19c, 20c or 21c - doesn’t actually matter except for the comfort of that room - keeps the rest of the house between 22c and 23c.  I do need to have 5 of 15 TRVs turned down and one replaced with a decorators cap and the landing radiator never comes on because it is perpetually at 23c - 23.5c.  It’s mostly good for me and the wife is very happy but it does mean that I have very little control over the heating - for example, I don’t really know how to implement a setback as only one room follows the thermostat setting.  In some respect, it may not matter as the heat pump is obviously working at whatever the set temp is regardless of how warm the rest of the house is: I’m paying for a 20c temperature but heating to 22c+!

I’ve nearly got the bedroom dialled in to a reasonable temp and once that happens I will try dropping the warm weather point to 35c and see the effect.  We’re rapidly running out of winter though.

I didn’t have the installation problems you experienced however I’m not convinced they have the heat loss correct even after the second attempt so I don’t want them dropping radiator sizes, getting it wrong and being in the opposite position with an unhappy wife!

Posted by: @andrewj

↑

My experience has been a bit problematic in that the radiators are mostly oversized and the flow temperatures can’t be reduced enoug

That statement in itself (which I am not in anyway criticising) exposes the limitations (I originally wrote madness, but relented because they doubtless have their reasons) on the settings that Octopus impose,  Radiators cant really be 'mostly oversized', in this circumstance a few are undersized, provided you can reduce the flow temperature!

That said its notable that, with the exception of the Cosy 12 issue, there are very few moans here (and lots of positive comments) about Octopus installs.  The reported SCOPs are also pretty good notwithstanding the firmware limitations, so they have clearly done something right in the design and implementation even if it could, in principle, be a bit better still with the right commissioning.  Their potentially great advantage (shared, so far as I know, only with Aira) is vertical integration which means that they can design the heat pump to deal with the practical limitations of the installer community, of heat loss surveys and of real houses and indeed of real customers.  Given that the vast majority of complaints we see here are down to the first two of these that should be a massive differentiator if they can exploit it which I would say, based on the limited evidence here, they quite possibly do.

@andrewj Have you considered just running it on a fixed flow temp? At the lower end there isn't really much to even compensate, your house must be well insulated. 

@jamespa Its a modern compressor with a decent driver, plus its R290! I have a feeling that they borrowed REDs nice dual stage injection technology, which injects refrigerant vapour to cool the compressor at high water flow temps. It certainly has lots of valves. Similar to mitsu's method.

Plus the integrated expansion vessel is great. I have also been very surprised with the performance, I just absolutely need the noise issue put to bed. 

However today I had a pretty funny experience, must have been the exact right conditions...

Went outside and the unit sounded like an absolute bag of spanners 🤣, whilst I was there it started its defrost and low and behold! A combo of very light, 3C almost freezing rain and I guess a low fan rpm had iced up the fan area. Resolved itself after the defrost, must of been like it for about 10 minutes max. I think there might be a FEA nerd shaking their fists saying "I told you fans at an angle was a bad idea!" 😆 

Posted by: @swwils

↑

Went outside and the unit sounded like an absolute bag of spanners , whilst I was there it started its defrost and low and behold! A combo of very light, 3C almost freezing rain and I guess a low fan rpm had iced up the fan area. Resolved itself after the defrost, must of been like it for about 10 minutes max. I think there might be a FEA nerd shaking their fists saying "I told you fans at an angle was a bad idea!"  

Wow, that sounds like fun.  There is another thread where someone with a Hitachi heat pump (featuring a more conventional vertical fan) describes a similar phenomenon.  In this case I have a suspicion that they have made the whole thing just a bit too compact, and as a result ice crystals can literally 'hit the fan' just like yours, but due to a slightly different cause.  My Vaillant machine has a gap of about 30cm between fan and evaporator, but the downside of that is that its 550mm deep in total whereas many Hitachi and LGs are typically only ~33cm deep.   Everything in life is a design compromise!

@swwils I believe it is well insulated: it’s a Finnish timber frame house designed to cope with Finnish winters.  The basic structure is 1980s technology, but the windows are new triple-glazed Internorm (Austrian) windows with a typical 0.7u value and the external doors are PassiveHaus standard.

I haven’t tried just a fixed flow but even so, I think I’ve come to three conclusions:

• the radiators have been selected to give different temps in every room (following MCS guidelines), albeit 10 out of 15 are oversized.  Unfortunately I haven’t been able to throttle any of them down far enough, yet, to cope with always on.
• the Cosy 9 is probably oversized for my house although I am still able to get good COPs out of it
• upstairs is a lost cause.  Having upped loft insulation to 300mm at the same time as putting in the heat pump and removing the cold water storage tank heated air is now trapped. The landing radiator never comes on because it’s perpetually too warm for it to do so.  Our bedroom takes 13-14 hours to drop from 22.5c to 21.3c with only an occasional barely perceptible warm radiator.

Currently we have temps typically around 22.5c and the wife is happy with that, and it’s achieving those temps by asking it to heat the house to 20c.  Having spent from mid-October fiddling with it to get it here, I just want to get on with my life and not worry about it anymore, at least for this year!

Interesting comment from Octopus’ Social Media manager yesterday on Facebook: the Cosy 9 CAN modulate below 30c and do so efficiently.  He always comes across as technically knowledgeable but that is a statement at odds with other Octopus technical employees, and also your experience IIRC.

I've got a fairly large volume system (12+ rads and a voulmiser, so probably well north of 100L) and it doesn't seem like my cosy 9 can maintain dT wide enough below about 32C flow, so it cuts off. So either my system is messed up, or the controls won't let it modulate below 30, OR 3000W is too much for my house even at 3C which seems unlikey.

Posted by: @andrewj

↑

Interesting comment from Octopus’ Social Media manager yesterday on Facebook: the Cosy 9 CAN modulate below 30c and do so efficiently.  He always comes across as technically knowledgeable but that is a statement at odds with other Octopus technical employees, and also your experience IIRC.

I have just read your very helpful discussion with TJ on FB and nearly fell off my chair, as he totally contradicts what he personally has advocated in recent weeks.  I am not sure where this leaves me as I am not now sure I can trust what he says, and he is the official face of Cosy after all.

Anyway, his challenge is to drop both ends of the WC by 3c and see what happens, so as I currently run at 37/50 I'll drop to 34/50 and check after 48 hours.  What I really want to do is to drop the cold end further, maybe to around 45c but that seems to be impossible...or is it?  Let's see what else comes out of the woodwork on FB.

@swwils What i deduce from all the comments in this forum is that hardly any installation has been setup. Mine by Octopus was just a simple spec, fit and forget. I had no instructions how to set it up, just a call us if there is a ln issue. It took a few hours before it fell over, leaving us with no heating over xmas period; just 2 puny fan heaters they provided to keep one room warm.

I had to spend 6 weeks arguing with Octopus that their system was not working; withholding final payment and a deadlock email from me was the only trigger for them to meet me at my house and discuss the issues.

I got 2 radiators changed and half the microbore pipes removed, also i pointed out that leaving dead ends causes turbulence in the water flow and consequently flow noise is higher.

I removed the remaining microbore and fixed the Towel Rail by removing the TRV. How they calculated a TR on full open will overheat a bathroom is beyond me.

The App allows you to change the min/max flow temps and turn off WC. Nowhere have I read turning of WC is a good idea; so WHY have it as an option In the App!

Posted by: @kevh

↑

....What I really want to do is to drop the cold end further, maybe to around 45c but that seems to be impossible...or is it? 

In the app, the minimum you can set it to is 50c but that may just flat-line the curve from the 50c point set in the hub to the -10C OAT mark.  What you can try is going into the actual Hub and finding the settings that give you access to your design temps - you may find that it is set to something like 50c at -3.4c (depends where you live).  You can change that temp - mine is now 50c at -10c but you can see that my house is prone to being very warm.  I've mapped various settings on a graph: to get 45c at, say -3, you'd want to set the 50c point to -10c (in the hub, not the app) and the warm end to 33c.  You can see from the graph, that the warm end adjustment can have a big impact on the curve, so just changing to 34c might be enough of an experiment.  In any case, only change one thing at a time and wait a few days in between.  And really, I feel it's too warm outside right now to make any real judgement, you really want to try this when temps are close to 0c or below.   

I've definitely reached the point where making micro-adjustments to achieve better SCOPs is wasting my life away.  I suspect that this was more a challenge to those who are running at temps like 57c to 38c and/or reporting SCOPs at M 3.5 rather than those who have already gone through the process of tuning.

Note that the graph is produced under "best reported internet knowledge" and may not be right - particularly in light of what TJ said last night in regards to the Cosy 9 modulation!  This one flat-lines at 35c.  I have made some statistically irrelevant checks against it and reported flow temp in the hub and it was only ever < 0.5c out so it's probably good enough.