TerryE

Spoiler alert: I have at last got my head around all this through a mix of simple analytic models and a simple 2D model of the slab. I wrote the model in Fortran to appeal to the other old farts that might just want to look at it or have a play. IMO, the bottom line is that using a buffer tank for a low-temperature UFH is entirely optional, though the controls and heating strategy are slightly different with and without. If you already have a buffer tank then it's not worth redoing the system, but if you are about to install one then IMO you can leave it out. (Note this primarily applies to an MBC / passive slab or equivalent house.) I've got friends coming around for a drink in 15 mins, so I will have to write this up later. But I'll dump a couple of graphs here for food for thought. The model outputs are entirely consistent with my analytic approximation. I will explain more when I next post. Until later ... (drinky time)

Well, that was a useful Q to ask, as the answer caught me by surprise. I don't suppose anyone knows one that they would recommend?

Gosh, isn't filtering the water supply a little extreme? I've managed 60-odd years without this. I've seen scale inhibitors installed. How common is this?

Just as quick Q, I came across this example in the Honeywell Pressure Reduction Valve (PRV) spec. I get the meter with isolation valves either side, the double check valve, the pressure gauge, the PRV anf the two main circuits to the right, but what is the device under the gauge? A scale inhibitor? Or any better guesses?

I've been looking into the options for meters and valves. These Solenoid valve WRAS approved 2/2 way Normally Closed 1/2"BSP are £56.41 ea. and I can live with the 100% mark up compared to JSHs suggestion for the WRAS label . The metering is more of an issue. I can get WRAS approved meters either with M-BUS or simple pulsed output. M-BUS is a PITA, but my main concern is that these are really for billing-type applications where you want confidence in total flow rates. But they can also cause around 0.2 bar head loss which is less acceptable. I might just slip in one of Neil's suggested Hall-effect flow meters which is really what I want anyway. If I do use this type of meter then I will have to have some manual bypass for at least one of the SunAmps so that I can at least draw hot water if there's been a power failure.

If you use the SunAmp inbuilt mixers then the full flow goes through the 15mm tails. By setting them to max, or 65°C say, then using the external mixers I will do a ~2:1 blend in the mixer with the DCW and so be able to sustain a higher flow. I also have 2 SunAmps in parallel and using the external mixer simplifies setup: only one TMV to trim. Anyway, that's my thinking, but we'll have to see in practice.

Nick swears by HEP2O. As to the layout, the basic rule if you use a manifold system is no buried joins if at all possible, and one (H +) C to one appliance. Our exceptions to this rule are: We have some elbows in the studding where the pipe is coming down the wall and needs to do a sharp 90° (e.g. the shower tails) We have the cold feed dishwasher T'ed off the kitchen sink cold. We have the cold feed washing machine T'ed off the utlity sink cold. IMO, laying out the pipe approaches to the manifolds is a bit like laying out a circuit board: it can get busy and messy and you need to avoid getting pipework in knots. Do your foulwater first because those are fat and difficult to route. Then the plumbing, then the electrics. I reserved the roof space in the GFL toilet adjacent to my manifolds for my "Clapham Junction" and actually did a full routing diagram for this before I laid a single pipe, and I had to go through three design iterations to get everything in the right order. Jan though I was being totally anal and going overboard -- until we started pulling the pipes and then she did a volte-face and agreed that it was a bloody good idea. Remember to space out the hot and cold manifolds to give yourself routing room. I am using 2×SunAmp instead of a TS or UVC. They immediately below the hot manifold so my hot copper runs total about 2m of pipework. That's it. Read through the various worked descriptions on the forum.

@jamiehamy, Have a look at the thread I've got running at the moment. We are doing ours in HEP2O using a manifold system as well. Deciding the runs and placing the manifolds is the bit that needs careful planning. We did all of ours in 15mm because our layout means that nearly all of the runs to the manifold were pretty short so standardising on one bore simplified it all. Though I admit using up a coil of 10mm for the low flow toilets and hand basins is something that we could have considered. I would question the need to use 22mm to the bath in a pressurised system. Unless you've a got ridiculously long run then 15mm will be fine, and avoid needing to use buried bends or joints.

@Nickfromwales, when I get rid of this sodding man-cold that Jan's mate -- well really our mate -- gave me, I'll be starting to pressure test our system, so I will -- for the first time -- have a quantitative measure of our mains pressure, but from the subjective thumb-on-faucet test, I reckon that it's above 2 bar and below 3 -- but I might be wrong. I'm a sceptic by nature which means that hard empirical data that I trust wins out over beliefs every day. However unlike May, Trump and those other wankers, if I don't trust the data then I double check it rather that ignoring or simply rejecting it. And you might have noticed that my preferred medicine for man-cold is legally distilled and purchasable. Ho hum. Sorry other readers. Thanks to the back story on our house, we have got 1×toilet, 3×wetroom en-suites, and 1× bathroom -- in a house that will have 3 relatively dirty occupants 95% of the time (that is we bathe or shower less than once a day on average). Given that I once had 3 sweaty teenagers in our current farmhouse with its tank-fed indirect system, our new house will have at least 2× the supply capacity with 60% the demand, so I really don't envisage any issues in practice. Over the last 6 years (when there's only been 3 occupants in the farmhouse) we have averaged about 270 ltr / day mains water draw-off for all three of us, so I don't see the need for huge accumulators, even if the mains pressure proves crap. My current thinking is that we do have a DCW accumulator, so that we don't have feed issues if we do have a couple of showers going at the same time, but something relatively small -- say a 70ltr accumulator -- which should carry us over the two shower hump. However, I will order the TMW(s). Ta muchly. Mrs E doesn't think that we need a separate drop-down for the low flow basins, but she's far more likely to value your views on this than mine, so post away with your pearls of wisdom. PS. Do you ever drive over the eastern border? If you ever have to go up the M1, there's a very cheap hostelry just off J15 (This offer is also open to other selected members).

So what is the current wisdom on TMVs? I am talking specifically about the mix-down of the ~ 55°C output from the SunAmp with room temp or colder Domestic Cold Water (DCW) to feed the DHW manifolds. Have you guys got any personal recommendations? Is 22 or 28 best here? The 28mm TMVs seem to cost crazy money. I've got 2 × 15 outputs from the SPV into 22 and the DCW into 22. The TMV output splits again into 2×22 a few cm after the TMV output. Our current house DHW is in a mix of 22 and 15mm copper which an internal cold header tank so is at ~ 0.2 bar and we've never had any problems. Is it worth a separate 15 or 22 TMV for the low flow hand-basins set at a non-scalding temp, say 48°C?

In 5 years time you might be saying "Oh, that's just so 2010s" and cringe. I think Ian would have apoplexy if you wanted to redo the kitchen in anything less than 10 years at that price

Why not just an accelerometer to detect the vibration / hum whilst running. Then you don't need to touch the mains voltage.

It's not good for your or your neighbours lungs either!!

Neil, that's brilliant. I am not looking for absolute total flow accuracy but flow rate that I can measure with an IoT device. This fits the bill exactly. I think that this level of instrumentation and automation is fast becoming a post go-live system upgrade. @JSHarris Jeremy. Thanks. Just what I was looking for. Much appreciate both your inputs.

You make a valid point This depends on the UFH flow rates. At a flow rate of 1 m/sec the UFH manifold will be circulating just under 30l/min. A full on shower might be half of this, so just doing the thermal balance at a 5°C to 21°C exchange, this would boost the supply to around 15°C. As you say nowhere near as good as 21°C but still worth having. Of course the efficiency would fall a lot if there were multiple demands, but this isn't a frequent pattern in our house and a lot of use: hand-basins, sinks, etc., are a lot less than 15l / min. And as I will discuss in a later post, there are times such as in the morning when the UFH loop will pull well over 25°C. This PHE might also be a post inspection update It really depends on the price of the PHE to see if doing this is worthwhile. @JSHarris Jeremy, what did you buy and how much did it cost? Another piss-ball is that the pulsed water meters typically generate one pulse per 10 litres which isn't really fine enough for what I want. Maybe I need a bit on LED sensor magic over on the face of the meter to directly read the little wheel which goes around once per litre.

@PeterW Peter, on-off is really what I want here. The two SunAmps are in parallel, mainly for capacity, but to a lesser extent for flow rate. What I am a little concerned about it balancing them, not so much in terms of max flow rate but in draw-down Joules. Maybe 95% of the time one SunAmp will happily feed the flow demand so my idea is that I always have at least one open (and normally only one, so that I don't get external heat losses from the other). I track the flow through it and the ∆T, so I can calculate the Joules draw down from it. I'll swap over once or a few times a day so that I roughly balance the use. Where we may need both open at once is on the very occasional time that 2+ showers / bath are going at the same time, and in which case I open both valves, and I assume a 50:50 split for draw down of Joules. I don't have photovoltaics (PV) so I only want to recharge them using E7 low tariff whenever practical. However, I also want to know (again less than once in a blue moon) when I am getting close to daily capacity and and here I will take the high-tariff hit for the convenience of not running out of hot water. @JSHarris Jeremy, I will come to the UFH on a post in a few days, but my issue here is that I am switching what the BRegs class as potable water, and I am putting this in before the B Inspector's final audit. So I need to show on my commissioning checklist and Bill of Materials (BOM) that all components have WRAS approval for potable water. He'll almost certainly not check, but he might because the setup is very different from routine and he might just be very interested in what I've done, and that's the rub. Of course the other option is to fit the valves post sign-off, but either way, I'll be occupant. Why the current BRegs stipulate that the hot supply must be potable seems strange to me, but they do . Maybe people like to gargle in the shower

So my specific Qs are: Any recommendations for the TMW, the 2 × electrically controlled shut -off valves and the digital / pulsed water meter, bearing in mind that these need WRAS approval for use on domestic cold potable water. I've got a box some 30×15×60 with all of my hot stuff in it that needs lagging. Any recommendations here? I was thinking of using that Aluminium covered quilting. Beg, beg

I am not sure whether this is best done as a blog post or a discussion thread, but since I am as much canvassing feedback and comment, I will start this as a thread and maybe summarise later in a blog post. I want tell a quick story as an introductory analogy. Back around the early 1980s when real computers took up most of a room, "Mini computers" were only the size of filing cabinet and ran at 1 MIP or thereabouts, and the first 5MHz IBM PC had only just been introduced, I remember reading a "futures" paper that took Moore's Law and extrapolated it into the future. What it concluded was that 1 GHz computers would be quite possible in the future, but identified that sticking point would be the memory: because of the speed of light, it would have to be less than 6" away from the processor! Unbelievable. Except that's how all processor boards are laid out today: memory on-chip or right next to it. My current house was designed around the time this paper was published: lots of copper piping and loops, the boiler on one wall, the hot tank 4m away in a cupboard, the kitchen 8m away again in the other direction. So the copper is in a mix of 28, 22 and 15 to get the flow rates -- and lots of copper to heat up on the way to the kitchen taps. Old mind-set. OK, plumbing technology hasn't changed nearly as much as computing, but a lot has changed, though many plumbers still hang onto techniques that were good practice 30-odd years ago. What I've tried to do in this new house is to abandon unnecessary concepts and optimise the system as much as possible for the use we want to put it to. So one core aspect is that we've already planned the house to centralise the wet works as much as practical: the house layout is roughly 3 × 2 tiles on two stories and a 3 × 1 in the warm roof. It is double fronted with living, dining and sleeping to the left and right. The front centre is a hall and landing column giving access to all rooms, and the rear centre contains the groundfloor (GFL) toilet, utility, bathroom, and one en-suite. Another en-suite backs onto it. We're using a HEP2O manifold system, so this means that all of the pipe runs are short, except the 7m run to the kitchen taps and a 6m run to the en-suite in the loft. The only long run of low-flow hot is to the loft en-suite hand basin, so we've just kept it simple and used 15mm HEP2O everywhere. What I want to talk about first in this post is our domestic hot water (DHW) system. It is heavily influenced by a thread where we discussed HEP2O manifold options with @Nickfromwales and others, and the example that he posted on this thread (with my current tails along side for comparison): This second photo is of my GFL toilet. The cold manifolds will be on the right (the two groups are high flow and low flow respectively) in a boxed in section above the Gerberit and the hot ones on the left in what will be the services area some 1.4m wide × 0.65m deep that will house all of our under floorheating (UFH) and waterworks and be separated from the toilet area itself by removable quick release panels. Some of my design drivers / rules are: All of the pipework in the service cupboard will be in copper, and sized 15 / 22 / 28 according to flow rates. My preferred approach is to use end-feed copper joints / fittings except where functional fittings are compression. This being said, I also use occasional compression joins to allow me to make up any soldered sub-assemblies on the bench, so that I don't have to solder in place. The hot runs have been reduced in size to the absolute minimum to avoid waste heating. I am using 2 × SunAmpPV (SAPV) for my DWH heat store. I am metering water flows and temperatures. I am preheating the cold feed for DHW to ~21°C using a plate heat exchanger (PHE) coupled to the UFH slab loops. So I have the 2 × SAPVs each using the supplied 15mm tails to a common 22mm @ ~55°C that is mixed with a 22mm @ ~21°C and fed directly into the DHW manifolds. The SAPVs are on a platform in the cupboard to keep this hot line short and to provide space underneath for the UFH subsystem. Each SAPV has a electronic value on its input feed, so that only one is normally connected in flow. Above a threshold flow rate, the second valve is opened to allow parallel flow from both; the "on-stream" and standby SAPVs are swapped during the day to balance the heat drawn down from the units. The PHE merits some explanation. In the winter our cold supply is typically at 5-7°C mixed which is the boosted to 48°C for supply to the hot taps. So that's a delta of 42°C, say. By using the slab to preheat the cold supply to ~21°C this drops the delta that's drawn from the SAPVs to 27°C and this effectively increases the supply capacity of the SAPVs by just over 50%. This is isn't free of course, because I will still need to add this extra heat to the slab on heating days, but we still effectively get 3 SAPVs capacity for the price of 2, and our maximum hot flow rate is prorated accordingly. On the thermal feed side, the PHE is (always) serially connected in the UFH circuit behind the return UFH manifold. The H/W flow rate is used to create a demand that will override the UFH circulation pump to the on mode. The intelligent charging and 1 / 2 use of the SPVs requires my Home Automation system to collect flow rate and temperature info from these feeds. My original plan was to put the digital meter and electronic valves on the hot output side of the SAPVs, until I had that "Durrrhhh" moment and realised that I could just as easily meter the input cold side. So here is the proposed schematic for my DHW. I have still to select the Digital / pulsed water meter, the valves and the Thermostatic Mixer Valve (TMV). But any personal recommendations / comments will be appreciated. The hot zone (that is pipework over ambient temp, and which must be lagged) comprises the 2 × 6 port manifolds, the TMV, some 1m of 22mm pipe and 1m of 15mm pipe a couple of full bore 15 mm isolation valves and the two flexible couplers to the SAPVs. Arguably, I should do the common section on the output of the TMV in 28mm, but given that its about 5cm long, there doesn't seem a lot of point. I haven't shown the SAPVs because they are just white boxes with some connectors at the back, that the flexible 15mm tails connect to. What I've missed off this diagram is the four temperature sensors: one on the combined input to the SAPVs, one on the combined hot O/P into the TMV and two one each of the output tails from the SAPVs.

Or just get one that doesn't have the glass in the first place. I think that Jøtul still do these.

The height is a trade-off: bending the back all of the time or lifting wood. I personally prefer using my legs to lift and keeping my back straight when I cut. As to the notch, IMO you should engage the dog teeth and then let the saw's own weight and its own pull rotate it into the log. No need to push, pull or force the blade down; if you are doing this then your technique is wrong. This being the case, the notch only needs to be half (maybe up to the full) maximum width that you want to cut. And keep your blade sharp. As I said before the files are cheap and YouTube gives you lessons on technique. If you do start hitting nails then it's probably eaiser just to order a new chain. They only cost ~£15, IIRC

Just make sure that you've got a full hot tank so you can have a good steep and top-up in the bath when you are done!

@Steptoe, the cock crows twice. Please leave it. You seem to be one vs the world on a typo. Have you never made one yourself?

@Declan52same chainsaw as me The maul set seems a bargain. Rob, the maul takes a bit of getting used to. Momentum and accuracy. Also the C of G is slightly off the axis of the shaft so the head naturally turns to split the wood on impact. I recommend making up your own bench out of offcuts. Then you won't cry when you accidentally saw it in half!!

Where you live, I think that we can give you a bye on going woodstove (And we do the same in our current farmhouse). Wood burning is high labour. There's no denying. If you want efficient burn then you need to cut and stack your wood for a couple of years to get the moisture content down (otherwise a large % of the heat energy goes into boiling off sap water up the chimney). If your logs need cut to length then you'll need a chainsaw, plus making up a logging cradle and PPE such as kelvar trousers, ear defenders and gloves. I can't remember how old you are but IIRC you are a couple of decades behind me. Even so, doing a 3-4 tonne load of logs is back-breaking work so you will need to be reasonably fit and do your yoga ( ... ) exercises to strengthen those muscles. IMO, your don't need a hydraulic splitter for green wood, though a couple of splitting wedges, a sledge hammer and a splitting axe are pretty essential. BTW, if you have electricity on-site then I'd recommend an electric chainsaw. I always had problems starting my bloody Stihl, so I swapped it with a friend and bought myself one of these. Bloody useless for wandering around a forest, but perfectly adequate for working in the garden at a saw bench, but buy yourself a set of sharpening files and watch the usual "how to sharpen a chainsaw" videos on YouTube.

I could reply or just to make it more fun you could read my blog and a few others such as Jeremy Harris. There's about have a dozen of the frequent posters who have used the same supplier without any material complaints or even reservations, but others have used other TF suppliers and different construction techniques. What we chose worked for us. It might work for you. What others have chosen might also work for you. But at the end of the day, you have to make your choice as to what is best for you. If you ask here then you'll invariably get at least 2 or 3 perspectives. What I really like about this site, is that I often get out of my comfort zone, so I can research what others have done or just ask the Q. In return I give back where I do have the knowledge to contribute.