ian192744

answering the HA integration: best way to do it with samsung is to install the modbus module MIM-B19 into the outdoor unit. Then modbus to HA machine (or modbus to an ESP, ESP to HA). I use this and it works well. My original intent was to use it for control and I do have the ability to do so, but I mainly use it for monitoring, I can see pretty much everything. for some idea what you can track: my system is https://heatpumpmonitor.org/system/view?id=45 on heatpumpmonitor. COP's are ok 4 ish in milder weather, but drop noticeably below 3 once temp is below 3C or so. Probably a newer on like the Vaillant would do better, if I was choosing again now that'd be on my shortlist. I can get my samsung 16 down to 4kw but thats pushing the lower boundary. I have a 200m2 1973 house, my house needs 10kw when its -3C, 4kw when its 7C . so its ok. but from what I've read on other fora, the 8 doesn't get down as well , can't do 25%, more like about 3.0 kw. If your house is 100m2 and well insulated 2019 new build then an 8kw samsung will be too big, its minimum of 3.0kw in mild weather will be way overpowered. make sure you research the minimum turn down of the options you are looking at. (most mfrs won't tell you, you need to glean from other users on fora). Are you able to carefully measure your current gas boiler use to workout your heat throughput?

just wanted to add my 2c to this. The figures in the Graham Hendra Freedom HeatPumps video here are the wrong way round. I recently self-installed samsung 16kw. followed grahams video's for setup. couldn't get my head round my radiators running cooler when we got a cold snap. the manuals are all very well but they don't say which numbered setting is on the right and which is on the left, and the screen doesn't say which one is which number. not ideal by samsung. anyway in 202* the "Low Target Value" is the lowest LWT temp value that you want to use when its mild . The "High Target Value" is the highest LWT you want to use when its cold.

near milton keynes. If there are companies that can do this, then any specifics please? yes have considered EWI but the house is quite an awkwards shape, quotes have been in the tens of thousands. so want to re-look at this first. also someone made the point to me that if the cavity has air in , then air will circulate in there, thus reducing the performance of EWI

Hi , My house is 1973 built, exterior skin brick, interior concrete block , cavity wall, detached. The cavity varies between 40mm and 80mm approx. I don't know why its not consistent or so small, it just is. The cavity contains 20mm of a very "old school" polystyrene board - somewhat like what would have been used to package a TV in 1970's or 1980's for those old enough to remember that far back. The type that fragmented into small white balls at the slightest poke. The board isn't even particularly well fitted - from the small bits of the cavity that I've had open at various time, its simply loosely laid in there not fixed. It is not a modern system at all. so the insulation is not brilliant, as you might expect. Every "bog standard" cavity wall company that I have had round over the years, after survey has said they won't touch it. All for understandable reasons - not enough remaining cavity gap to work with , danger of getting cold spots with an incomplete fill, risk of the existing polystyrene board fragmenting etc. I'd sort of given up on it but now that I am trying again to improve the fabric, its the first thing again that comes to mind again. Can anyone point me at a company that might have a solution for this? thanks Ian

thanks. sorry but your diagram loses all detail with the screenshot resolution. however, I have studied dozens of similar diagrams , plumbing and heating websites, watched umpteen design videos, and it does seem to be common that the pump for the secondary circuit is shown adjacent to the buffer pumping away , but nowhere can I find a reference that says that this is the right thing to do. DHW is in a 1 year old 300L UVC (which nick steered me towards after a previous discussion a year or two back) and has a HP size coil so not changing that. the buffers I've been looking at are this https://lovatospa.com/gb/product/49017080/bomber-20-hybrid to fit without difficulty, or possibly, with some inconvenience and significant re-arrangement of the airing cupboard, this https://lovatospa.com/gb/product/49017084/bomber-50-c (which does have an immersion option). I REALLY want one that will flat-to-wall mount because of where it has to go. so many uk supplied buffers seem to be just cylinders and thus take up a lot of space projecting outwards. great if you have a plant room, not so great if you don't. boiler is a heat only, no pump, hence the question. Currently pump is away from boiler into rad circuit (at the mid point).

+info wouldn't let me edit above post.. too slow ... MCS heat loss for my house at -3C is 14-16kw depending on room temp assumptions. Boiler has been flow-temp limited to 55 for the last year and half. gas load monitored on 30 minute intervals, I know we're comfortably inside 16kw almost all the time.

hi nick The e-boiler is a backup for if the rest is broken, just heat a couple of rooms . Many ashp have a backup heater - I'd rather get one without a backup heater and have the backup contained in a separate device (so its a proper backup!) . Its also to heat 1 room regularly . the house has just 1 room occupied during the day 4 days a week - just my study - and I detest having to run the gas to heat this when I could use electricity which is either low-carbon (Tou into battery ) or from PV, so cold sunny days in feb-march-april it will be free. none of this is predicated on cost saving with ToU tariffs (which I happen to believe will continue, but thats by the by) just a nice to have. why would a 20L buffer be unsuitable for this design? not big enough? there are 16kw heat pumps (e.g daikin) that are specced for a min vol of 20L which is why I am at that figure, its not a made up number. 100L would be really difficult to find a space for. there is no intent to run heat sources simultaneously. the heat pump in the diagram doesn't exist and won't for some time.A2A non-starter - whole house is rads. This isn't a major rebuilding project. any thoughts on the actual hydronics question or do I need to look elsewhere for that? cheers Ian

hi all interested in the hydronics wisdom with regarding to pumping direction relative to an LLH / small buffer. Some system components like ASHP have their own pump so that is a given, it has to at the far point away from the LLH. however for a loop that needs its own pump (rads, and a heat only boiler), the pump needs to be sited physically close to the LLH in my case - there is no where else it can go. Should it pump away from the LLH (positive pressure into the circuit - pumps A and C ) or towards the LLH (positive pressure towards the LLH - pumps B and D). I'd have expected from the physics of it that you'd want the positive pressure going into the circuit , but struggling to find any decent reference documentation on this? reasons for system design - existing system is not great . looked at it all sorts of ways, we have a flow rate problem. all the loops are long and bendy, big house, lots of rads, TRV's on all. boiler short cycles a lot, and in fact has been known to not fire due to not enough flow rate when TRV' drop down. separating primary from secondary will give full flow to boiler, and allow pump tuning for the needs of that loop, rather than the current which is a compromise. - will put in an e-boiler (as backup for the gas) asap. cheap and cheerful , doddle to install, and running on 40kwh of octopus go stored in my 3x powerwalls its a wee bit cheaper than gas. - intend to add an ASHP next year but will keep boiler. not been able to pin down a vendor or installer yet . capex for a 16kw + install struggling to make sense of right now. but 20L LLH buffer future proofs it. looking at a buffer with internal baffles / directed take offs to maintain stratification ta Ian

did either of you solve this question? I have honeywell evohome with opentherm to mod-con viessman boiler. looking for an ASHP that can consume third party stat input (large house, need programmable TRV's for reasons much discussed already, I know I'll need a fair-size buffer to maintain volume when TRV's shut). if ASHP can consume opentherm instead of on/off, so much the better.

planning my ASHP location, and pipe route to get to where all the rad circuits distribute from and likely buffer location - 1st floor airing cupboard. seems to me like my easiest option will be to go outside , up from ground level , round an awkward corner, then up the side of the house to roof level (2 story house), then into the loft space, across the loft then drop into the airing cupboard. If I go for the option of "get inside as soon as poss", the internal route will require a whole lot of lifting of boards (chipboard - 1970's house so a pain) , lots of joist drilling / notching and we have to go through 2 external walls (there's an extension that causes this to be 2 not 1) - all quite a lot of hassle and time consuming. Plus with 32mm pex, thats big joist holes. Prefer not 28mm copper as seriously pricy these days. whereas the external route for my house I reckon is do-able really quickly , mostly DIY able , and little/no major hole drilling. once into the loft, I can bury pipe under insulation and the routing is easy. Question - has anyone used pipe like this https://ingoodnic.uk/25-25-90mm-2.html ? I'm thinking 10m of dual 32mm preinsulated as 32mm has more or less equivalent ID to 28mm copper which seems to be the recommended by most installer. It'd go the side of the house and in through a hole in soffit into the loft space. No wall drilling. My soffits are really wide so room for bend radius should not be an issue. It'd be in a non-visible corner so no issues with spoiling appearance or similar. ta Ian

there's absolutely no vibration. I bought a 100pack of these guys https://www.ebay.co.uk/itm/144017582577, as long as the rad is the right shape to get a good contact, they won't move. PWM allows you complete flexibility with fan speed if you're unlucky enough to find a resonance at some point. But a bit of neoprene superglued around the contact points would be easy peasy to add if vibration damping was needed. I used to have a bit of an obsession with SilentPC's, I'd always build my own in the days when aircooling was the only option. I now buy fully passive heatpipe+heatsink and/or All-in-One water coolers for PC's that I uses, but its given me the background on building for airflow and near-silence at the same time. Mid-spec PC fans are really quiet and don't cost the earth. The arctic PWM PST fans are so useful because they come with the daisy chain connectors as part of the design - so only one upstream connection for as many fans as you like. I was thinking about top vs bottom of rad as mount point for cooling as well, although I don't have a cooling capable source yet. One bonus of magnet-held setups is that you do have the option to simply move it from the top of the rad to the bottom or vice versa. rads with less than 80mm depth are indeed a problem. depending on what else is below the rad you could mount an 80mm fan on an angle (lower end toward the wall, some trickery with clips into the skirting to hold that side up). I tried this with a K1 rad and it was ok but not great, obviously significant airflow wasted going behind the rad into the unheated space. I didn't experiment with blocking it. unfortunately I haven't found any 60mm fans worth trying - of course they exist, but they are all designed for comms room / datacentre environments where low noise is not a thing - plus to cover the length of the rad, you need loads. I think 80mm is the smallest worthwhile. I think something like this https://www.amazon.co.uk/Myhomeware-Radiator-Adapter-Diverter-Booster/dp/B07LGCXQZV?th=1 should be easy enough to make , in order to divert air away from curtains at the top, and if reversed with fans blowing down from the top and mounted underneath, could probably form the basis of a condensation drip catcher for cooling. I haven't done anything about that yet though. the other thing to think about is that if you CAN fit a K3 rad in for depth, then you'd have space to use 140mm fans which will have loads more airflow and you need less to cover the length - you could then make the rad smaller in either height or length. basically I am trying to avoid having to replace all my rads, especially not put monster K3's in everywhere. I'd rather have a K2 with a fan. having looked at the design of some of the commercially available FCU's , I couldn't find any where I liked the way they worked (fans too small, not controllable enough, no noise specs on the actual fans themselves - only on the overall unit) and/or pricing levels are very high compared to that of a rad plus some fans. And being a bit of a tinkerer I thought I'd build my own! I'll be building the mk2 at some point soon, I'll take some more detailed pics. Ian

this answer your question? obviously its all off right now....

I looked at the commercial units and decided a) too small b) no control over the quality or noise level of the fans used (undoubtedly cheap crap) c) quite pricey for what they are so I'm building my own. Fitted unmounted controller ler fan unit components: Arctic F9 or F8 PWM PST Fans (daisychained) Cable ties to hold the fans together Circular Neodymium magnets with countersunk holes, Bolts to hold them on Magnets itself to the underside of a flat panel K2 rad very nicely. Not nice enough to be commercially sold obviously , or to be on display to the world, but if its out of the way or you're not fussed about elegance, its ok. control unit: 12V PSU ESP32 controller , PWM output for fan speed, 3.3v (ESP) ->5v (FAN PWM) level shifting . Using a board from https://www.mottramlabs.com/index.html 12V relay for fan power on/off DS18B20 temperature sensor (gaffer taped to back of radiator) All run from ESPHOME and HOMEASSISTANT Photos are of the prototype version, tested end of 2022 winter. Temp sensor turns on the fans when rad is over 30C, ESP32 PWM controls the fan speed between 20% and 35% up to 50C flow. Its silent, and generates a noticeable improvement in room warmup time (I have not measured scientifically just by the "does Mrs Ian put more jumpers on or not" test....) I currently have a gas boiler was running it at 50-55C flow (with heating only of course!) whilst testing this in preparation for ASHP hopefully later this year. Ian

so at 26mm ID its the same bore as 28mm copper so flow rate calculators that assume 28mm copper should be valid. And the compression fittings make it do-able with a spanner. that is really excellent thank you! presume the crimp fittings need the right crimp tool (that your plumber had?), looks like they are about 100 quid, worth having if you are doing lots I guess.

brilliant thanks jenni. that is way cheaper than any copper solution. Do you know what the inner bore size is for that pipe? What fittings do you use to reduce it to 28mm to connect to the rest of your system?

looking at exactly that kind of thing. I've read previous posts on here mentioning the use of "32mm pipe" for HP projects but without being specific about what they used. really hoping someone will come along who said "we used <reference> and it works!"

that is exactly my plan. except that I am not convinced that 28mm (hep2o or john guest type) is big enough.

I wsa reading the samsung 16kw data sheet. 12l/min minimum, 58/l min maximum. Given that we want to be able to run it at full chat in worst, surely designing the maximum is the sizing number? agree with you , probably will need a bigger pump and am allowing for that. however what I don't want is to have to run a pump (or two) with very high head and high power due to noise and efficiency. would rather put in a bigger pipe! do you know of a better online calculator? I do have an engineering degree but that was 25 years ago and I went into IT so all the fluid dynamics stuff is totally gone...

I have to have a long run somewhere. If I put the buffer next to the ASHP then I'd have 15 metres from the buffer to the rad distribution point. so I've still got to move the heat to the rad distribution point over that distance, at a lower temp, so need higher flow rate . albeit that the flow rate of the HP would now satisfied by the buffer. I also know that the rad distribution is quite complex as well - so if I can split the circuit in half and make one half of it "clean" then I'll be gaining. lastly, in my case putting the buffer next to the ASHP is not that easy in terms of the physicals of the rest of the property.

Hi all I have a long convoluted primary pipe run from boiler location to airing cupboard , where the distribution to radiators occurs. Logically the change to ASHP best works with the ASHP outside where the boiler is, buffer tank in the airing cupboard where the radiator distribution is. A lot of the existing primary is 22mm, and it has loads of elbows (7-10 in each direction) . I think its about 30m its "evolved" over time due to various incremental changes in the heating setup of the house. I cannot get over 0.8m3/hr (13l/min) on it regardless of using pump speed III, number of rads open etc. Existing pump is a grundfos alpha2 15-50/60. various ASHP fitters who have visited are all sucking their teeth at the problem, suggesting pumping in both directions or very high heads or "it has to be 28mm". Even with 28mm I don't think it works. certainly not if using plastic. however, I can find an new primary route which will be about 12m each way with 3-5 elbows, and the potential to make some of them swept bends if in plastic. Call it 35m in total from HP->Buffer->HP. I am likely to need a 14-16kw heatpump. so I'm using 58l/min as my flow target for a samsung 16kw. If I get away with less, great. using https://heatpumps.co.uk/pressure+flow/simulator.html to do the maths. 58 l / min over 35m of pipe, 20% glycol, 35mm copper => head of 2metres -> easy for pump switch to 28mm plastic (22.4m ID) we only get 34l min with the max of 5M head. switch 28mm copper needs head of 4.8m to attain 58l/min with same .. just about ok. 35mm copper is of course ridiculously expensive and hard to work with. particularly as I intend to lay this run myself to save cost (I have some spare time). 28mm copper still not great from a cost or working on it point of view. bottom line question - why I am here : is there a plastic pipe size and type suitable for CH primaries that is greater than 28mm, available from UK merchants? Obviously also one that you can get elbow, coupling and reducing (to 28mm at the ends) fittings for? any specific references appreciated! ta Ian

room size 7.5M x 2.5M. ceiling height 2.3M (external side) vaulted to 4.3M (house side). 2 short and 1 long walls external . external area = 22m2 walls , roof 24m2 = total external area 46m2 of which glazed area (2x1.8) + (1.5x2.25) +(1.7x1.3) in walls + 2x(0.9x1.0) in roof = 11m2 I have 10 linear metres of regency OG thermaskirt on 1 short and 1 long side. Ian

I have discrete heat thermaskirt skirting board in one room. I went for one of the taller profiles to get the higher output . Main reason for doing so was there is literally nowhere in the room that rads can go without utterly ruining the room usage / layout, and we have no UFH anywhere so adding UFH to the system for just one room would have been overkill. Its a ground floor "garden room" which was a refit of an existing extension. Its uninsulated concrete floor, solid wood flooring on top of that, with two large windows, a patio door, and two large openings into adjacent rooms, one open plan. what relatively little wall there is, needs to have furniture against it. It runs on a gas condensing modulating boiler at CH temps as separate zone in the CH. It works ok, but due to the absence of convection it warms the room much more slowly than a rad would, and In the coldest part of winter it doesn't have quite enough heat output to keep the room cosy - because the room has a lot of glass its heat loss is pretty high. the room-based zoning (evohome) is therefore key to not wasting heat on an lossy room when its unused . If I had to chose again for this room I would probably end up the same due to all the constraints, but if any other room where the option to use conventional radiators was not impossible, I'd stick with conventional radiators. If the rooms in your house are better insulated than the particular one in my house, then you might well be ok with the output level. I fitted it DIY with labour assistance from the builder who was doing the extension refit work (some long sections, 2 man job). I also had the copper fin type skirting heating in my sitting room - original from the house build (1970's). it was only along a relatively short section of wall, not including the coldest one, and it wouldn't keep the room warm. I ripped it out and put two conventional K2 rads in instead - as there is plenty of space for rads in that room. now its warm! ta Ian

hi nick thanks for your advice and time. I was already seriously considering the 300L UVC option - certainly if I was doing this from scratch with no existing system fitted, thats pretty obviously the best option, have read lots of your previous commentary about that. However, ditching the TS is a big thing - both in terms of physical work , cost, and mental commitment (SWMBO view of nothing wrong with it , its paid for, why get rid?). would adding the buffer be a good thing to do, regardless of any potential decision about ditching the TS to put in an UVC? The objectives being - separate the flow speed , flow temperature and demand curve/timings on the boiler loop from that on the CH loop - facilitate the future ASHP (which is nothing more than an idea right now, probably 2-3 years off). immersion duress boost in winter isn't that sensible - it will draw from battery (which has charge overnight on octopus go) so although only charged at 5p a unit, its a cycle through the battery that is not best use of its lifetime cycles. I was thinking instead I could run the system boiler through an external plate hex, creating a pseudo-combi for the duress boost situation. That gives an immediate lifestyle benefit of enough hot water at all times in winter! See new diagram below. shopping list would be 2x TMV, 1x plate heat exchanger, 1x TDV, 1x pump, 1x 3 port multiposition valve, 1x 100L buffer so nothing too major. does upping to an 8m head CH pump make sense? the PP3 curve is steeper on the 15-80 , so am I correct to think should get more flow rate at the same radiator resistance/pressure level when all are open circuit? regards Ian

hi all long term lurker and reader on here, this is my first post, hopefully I follow the etiquette! House is 1973 built 2 storey 4 bed detached, brick and block, "too-thin" 40mm cavity wall with 20mm EPS in from build time, cannot add more due to lack of width. Loft has 400mm wool DIY'd. Floors are uninsulated concrete. double glazed. All rooms have large radiators, 20 at the last count. flow temps of 50-55 warm the house perfectly well. EPC says 22MwH, in practice we use 15MwH per annum .Full Honeywell Evohome control of CH and DHW. Gas boiler Viessmann 28kw 100WB1B 4 years old. DHW is in 1st floor airing cupboard - a DPS Pandora 210L open-vented store with solar thermal coil at the bottom (which is now unused ). DPS and solar thermal installed 15 years ago. thermal system has gone - it died (long story, not for here) then replaced the roof use with 12kw of PV.. Mains gas, Mains hot water (good pressure 5 bar). 2x Tesla powerwall2 batteries. problems to solve: 1. need more hot water capacity in winter : Condensing boiler cannot achieve the designed store temp (75c). It worked fine with the old nasty non-condensing one!. Boiler has to be made to run extra-hot (79C flow) to reheat store, but even then - 67C is hottest store temp that boiler can reliably achieve in a reasonable amount of time. Bottom part of store is unused capacity. 2. DHW reheat necessitates CH shutdown for 30 mins always happens once , often twice an evening during peak childrens shower time in winter. takes boiler 5 -10 mins to ramp from 50 (CH flow) to 79 (DHW flow target) + another 20 to reheat. if not CH not shutdown , the CH cold return doesn't let the boiler get hot enough, so the HW reheat takes ages. Plus the rads overheat due to the too-high flow temp. 3. Rads group 3 are poorly/not heated if groups1 and 2 are calling in large numbers. Pump is in PP3 mode. Group3 is a long convoluted pipe run , can't do anything about it. Pump on fixed speed3 will reach group3 but is very noisy (very loud resonances as well as flow noise)- when load is low (1 rad open a tiny bit) so have to use a PP mode 4. shoulder season short cycling lowest boiler modulation is 7.5kw. Load with only 1 or 2 rads , house in daytime occupancy, is way below that. 5. increase hot water storage capacity spring<->autumn: PV dump heats upper store to 90C easily - immersion is between the upper and lower coils. increase storage capacity by moving heat into bottom part of store 6. Add ASHP in future, without changing everything else. ASHP would be sized down so still use the gas for the worst days. existing system as per diagram1 constraints - no space to add any additional large tanks or stores. Loft is impractical (tiny hatch, no space for a big tank). No ground floor space that works . Don't want to throw away perfectly good existing kit (store and boiler). my thinking on solutions diagram2: problem1+2 -> loop gas DHW mode additionally through solar coil . Heat in "base" mode through on lower coil only whilst keep CH active. diagram3: problem5 ->add de-stratification loop to run primary water in a loop between the coils until temperature evens out diagrams2 and 3 aren't an either/or they are logically sequential steps which would in practice be implemented at the same time. problem 3-> increase to a higher head pump . current is alpha2 15-50/60. If upgrade to the 80 will this help as its PP3 curve is steeper? diagram 4: problems 4, 6(+3) : rework to include a100L sealed system buffer tank (such as this) to separate CH and boiler flow sides . Mount it ABOVE the DHW cylinder - There is a 3 sided internal wall recess to fit into, so strength to build appropriate wall-mounted supporting frame for it is not an issue. Set a higher flow rate on boiler side vs CH side to decrease store/HW reheat times. Can ASHP and boiler share same primary tappings into buffer via Tee, or better to have additional tappings put in? Not shown on the diagram, but also wondering to perhaps take from the buffer via another plate HEX as a "winter preheat" for the HW, to avoid the cold water being out-of-the-ground cold when it hits the DHW store, to further improve DHW delivery capacity? I'm hoping I haven't completely gone down the wrong track here - not spent any money on new stuff yet but quite some time thinking about it and analysing the behaviours with temperature monitors etc - would appreciate any advice anyone can give. thanks in advance Ian