TerryE

Lots of history and debate on other threads about this topic; have a search. Wood burners are highly problematic if you have a high energy spec house with MVHR. A typical wood burner might have a minimum 4kW output and dumping this amount of heat into a single room can cause problems. We have UFH which typically comes on overnight for however long needed to do the daily top-up. Used to have wood stoves in our old farmhouse but don't miss them or the dirt or need to feed them in the new. We don't have any heating on the top 2 floors, but do use a small oil-filled electrical heater on the first floor for a few hours top up overnight in the coldest months. You need to be careful in designing out edge thermal bridging in your slab profile. An approximation. It averages to whatever your day-round year-round average temperature is in your locale. In our case our slab is at ~22½°C year round and even with 300mm EPS, this heat is constantly bleeding into the ground under the slab, so I suspect that ours might be nearer 10°C.

Sorry to be late to this discussion. As you know I have a PHE in parallel to my single Willis, and these are configured in a single loop with my UFH manfold feeds to 3 underfloor loops that are all ~100m. This is a closed system with its own 5L expansion. Can I first suggest that you discount scaling. Surely yours is also a closed system? In my case I've got maybe 40L of water in my system. Top up is manual and I haven't had to do this in ~2½ years of running; still at the 1 bar that I filled it to at commissioning. How much hardness is there in 40L of water? As I've discussed before I have a lot of temperature logging -- 12 digital thermostat sensors that I collect and log every 2 mins since commissioning. My pump is on a medium setting and the single Willis at this flow rate lifts the circulating temperature by roughly 7½ °C though this is mixed with the PHE bypass dropping the overall mixed to around +5°C. So last night the heating came on at around 3:02 AM with the return starting at around 22½°C with the out quickly climbing to 27½°C. This 3kW being dumped into the slab steadily lifted the return temperature to 25.7 °C at 06:38 when the heating was turned off. At each midnight my control system calculates the top up needed for the coming day based on the forecast average external temperature with an adjustment based on the 24hr average house temperature against setpoint. This ~3½ × 3 kWh was what it calculated as needed. Note that my (NodeRED) heating control turns the Willis off at 35°C as a safety limit (keeping the pump circulating) though IIRC the hottest out flow that that I've recorded was just over 31°C. I have my manifold TMV set at 40°C (mainly as a safety backstop) so the valve is in reality always open and any output from the willis is always dumped into the slab. I note that you don't seem to implement any temperature control on your Willis outputs, yet have your TMV set at 35°C. So what happens if your loops cross this threshold? The mixer will start bypassing the UFH loops and now the output from the Willis is on a short closed cycle isolated from the circulating pump and with no heat dump and so you are heating maybe 3L of water with 6kW. No wonder it starts to kettle! IMO, the thermostat cutout on both Willis heaters should be set to a maximum of 35°C (or as in my case I have them in-built Willis sensors at minimum and I use a digital thermostat on the Willis casing for on-off control) and at least say 5°C less than the manifold TMV to ensure that you never short cycle through the Willis heaters. You can't dump 6kW × 24hrs into slab without seriously stressing something. You have to validate your design and assumptions during commissioning otherwise you risk permanent damage. At least with a TMW set at 35°C any damage will be in above ground replaceable components. PS. Adding extra short circuits is only going to make things worse.

We heat overnight so ours cycles between 21 and 22°C. We don't like it any cooler, but our kids complain when they visit.

It's not paint. It looks more like penetrating damp either through from the neighbours side or through a breech in the DPC. Notice how it looks worse near the floor. This is easy to confirm with a damp meter. No quick fix, that I know off.

Yes, you can be screwed if your slab is off-spec, but one way to mitigate this is (as we did) to get your slab and TF from the same supplier. I also did a full dimensions and levels check after the slab was finished and before confirming delivery of the frame. In our case we had a slump of about 3mm in one room which we decided to accept, but the frame plate base was true and accurate so we were happy to proceed. IMO, this detailed check is essential. We don't notice intra-floor noise, and a twin-wall frame filled with pumped cellulose filler is amazingly solid. A friend has a PUR insulated single wall TF; this has a smaller cross-section, but isn't nearly so substantial, IMO. Between floor noise is more of an issue and we do notice this between my son's bedsit on the top floor and the main guest room below on the 1st floor. However this is more a consequence of using eco-joists than the TF itself. This being said, EcoJoists make installing all of the between floor services such a doddle that I would still use them if I were doing this again - but I would install acoustic decoupling in the guestroom ceiling. We have a pretty high spec in terms of energy efficiency, so we only have UFH on the ground floor, and non on the top two floors. In the worst couple of months in the winter, I do run a small 2kW heater a few hours a night in my study / 2nd guestroom with the door ajar and this keeps the hall space at around 21°C. We prefer the the bedrooms being a degree or 2 cooler.

@Robert Clark, If you are talking about requiring bottom openings then I take it that your plumbing comes up from the floor. In this case having pipes on the surface going into a semi pedestal is going to end up looking really tacky, IMO. I can see only three options here: A full pedestal which hides the pipework. Note that Grohe and some other supplier do a variant of the basin + semi-pedestal where the basin is wall hung and not really carried by the pedestal. The sink can be fitted and plumbed up then the full pedestal fitted in place to hide the gubbins. Some form of vanity (or half vanity) unit Battening out the wall or even just a 1m wide vertical section with a 35mm stand-off and extra plasterboard layer. This does complicate tiling slightly, but can look quite striking especially if you use a contrast tiling scheme to make a feature of the standoff. If you do this, then remember to allow heavy duty supports behind the sink line to allow the load bearing. The pipework can be run up the wall and up to semi-pedestal levels. A lot cleaner IMO than a vanity unit.

We've got two; neither open bottom. One is a Grohe, IIRC. They have a couple of special brackets which you screw to the back wall using a template. You then position the semi-pedestal under the basin and a couple of fixings screw through the pedestal into the wall brackets holding the whole thing in position. These have colour matching covers so the whole thing is very neat. We've had no issues with either.

No need to calculated cepstrums and try to calculated estimate impulse responses or the like. ? The thermal inertias in a passive-class house like mine as so high that a simple liner fit to historic data is enough to estimate the basic constants. The calcs are run once a night and decide "tomorrow we need 3½ hrs heat into slab" or whatever. Magic happens, and we never really think about it. I used to plot all sorts of plots and analyses, but I never bother these days. The house just sits at the temperature we like.

My heating system calculation is pretty simple in essence. It is driven by three measures: The delta between the house target temperature and the outside temperature averaged over the day The day-averaged actual temperature delta'ed against the target temperature. These two are calculated each midnight. For the actual temperature, I have a DS18B20 buried inside an internal wall (actually the one my control system back onto so it was easy to drill a hole into the plaster and drop one down into the studding). IMO, doing this is a lot better than measuring air temperature as it gives a better estimate of the overall temperature of the internal fabric. For the outside temperature, I actually use an API that the Met office publishes and query the outside forecast for the coming day. We don't have any funny microclimate issues and this is good enough for what I need. Another option that I have considered is that our outsidel meter box is flush-mounted in our external stone skin; I have thought about dropping a DS18B20 down the wall void behind this between the outer skin and the inner lagged Larson strut timber frame of the house, as this will smooth out any spot surface solar heating noise. So each night at midnight I use these two terms in a simple linear model (i) to calculate the amount of heat in kWh that I need for the next day, (ii) to uplift or drop this to adjust if the house is too warm or too cold. I set the two constants in this calculation initially from my Jeremy-like heat calcs for the house, but tweaked them after a year by a fit to the actual house data. So what this daily calc does is to work out that I need to put, say, 12 kWh heat into the slab tomorrow to keep the temperature of the house on target. Given that I still haven't bothered to install an ASHP but use a 3kW heater into the slab, this equates to 4 hrs heating is needed so I heat the slab from 3-7a.m. (to use E7 low rate electricity). The last measure is the actual house temperature. If I need any more than 21 kWh, then I add this first 21 kWh midnight to 7a.m.; I only add the extra heat drip-feed N mins per hour if the house temperature drops below the target temperature. Because all (or sometimes most) of the heat is added overnight, there is a residual <1°C daily ripple on the actual temperature, but we find that we don't really notice this. The scheme works really well and the day-to-day average temperature varies maybe ¼°C. The one thing that does kick the temperature off is having visitors to stay because these extra warm bodies aren't factored into the heat calcs. This system runs for about 6 months a year, but cuts out for the summer hump where we don't need any heating. Note that we have smallish cottage-style windows so we don't have solar-gain management issues. As to slab temperature, we run our pump 6 mins every hour when the heating isn't on, just to redistribute any local solar gain across the entire slab. I take the average temperature of the flow returns coming out of the slab at the end of this 6 min period and this is a very accurate measure of the average slab temperature. No other probes needed. We already have a ~300m long probe.

I've got an energy efficient house with a comparable but perhaps slightly lower spec than Jeremy's. If you start with energy efficiency as a design goal then it doesn't cost that much extra than a normal BReg compliant build and there is a huge benefit in the quality of the lived in environment: air quality, total comfortable thermal environment, and the savings in running costs of course. These costs were lost in the noise compared to some of the extras that or planners forced on us.

We've got our wall cassettes and flooring on 400mm centres. Everything is just that bit more solid, IMO, compared to friend's house where he had 600mm centres.

@DavidHughes, you seem to be adopting a very robust "prove my figures wrong or I'll ignore you" based on a set of predictive calcs. I did my calcs 3-4 years ago and came to different conclusion: MVHR would save money and materially reduces running costs. After almost 2½ years of live data and analysis of running costs, our actual energy bills and house performance closely tramline those pre-build calcs. I can't recall the exact numbers but IIRC our system cost less than £3K to install and the heat recovery tipped us into the domain where we didn't need any space heating on our 1st or 2nd floors, so saved us a shed load of CH costs. We keep our entire house at 22-23°C 24×7 and our total annual energy bill is about 60% of our rates -- and that is with electric element heaters and not ASHP heating. Our as-built experience is that MVHR saves money and materially reduces running costs. Everyone that visits or stays at our house notices how fresh it feels / smells.

@MikeSharp01 The thing that I like about the ESP Tasmota or Lua Firmware -- and the Ardiuno sketches for that matter -- is that the H/W libraries just seem to work. Hooking an ESP or an Ardiuno onto an RPi via the USB serial interface is simple and has no external dependencies. It just works and I can (and currently do) include simple low level watchdog timers so the system is failsafe: if the RPi dies or disconnects then at worst the house will drop a degree a day or so until this is picked up. No heating circuits fail on. The raspberrypi.org documentation on how to use the RPi GPIOs is totally inadequate; you have to mine community resources such as the Adafruit tutorials. I was talking this through with Andrew earlier and in fact he is more comfortable with using the ESPs since he has had some trouble in the past configuring the RPi GPIOs for control use. He wasn't too bothered about the custom firmware on the ESPs so long I held a couple of ESP modules pre-imaged with the firmware sand held as cold swap spares. In terms of the physical configuration, the ESP setup is neater IMO: I have I/O front end in an ABS box that is connected to the RPi by a USB connector. This ABS box has all of the relays, flow meters, and DS18B20s plugging into connectors on it, and yes it also happens to have an ESP module inside. If I get rid of the ESP with its USB connection then I will need a custom connector containing at least GND + 3.3V + 5V + 7 I/Os connecting to the RPi GPIO header. If I go the ESP route then I can avoid Python altogether and just use NodeRED. So more thought needed.

Background I've blogged and posted separately about how I currently control my Underfloor Heating (UFH) and Hot Water (HW) systems using a Raspberry Pi (RPi) computer. This takes input from some digital thermometers, and a couple of digital flow meters (mainly for flow logging) as well as Met office weather forecast data to calculate timings and to switch the power circuits for my Willis heater, the circulating pump and the supply to my two SunAmp PV units. The initial design choices in my current implementation were to minimise my own learning curve and to mitigate the development risks for me. For example, I was already one of main the developers of an ESP firmware platform, so it was just easier for me to use a couple of ESP modules to handle all of the external sensor and relay control. That being said, this latest COVID-19 outbreak underlines our need to have a practical fallback for Jan. As it happens, both my son and son-in-law work in the IT industry, and both use RPi based home automation systems in their own homes, so either could in principle take over maintenance of our RPi-based system if necessary. Even so, on reflection my system setup is more complicated than it needs to be. In particular neither Rob nor Andrew is familiar with the ESP module firmware and development system, so this subsystem is not something that I could easily hand over. Given that a single RPi could do the entire computer lot quite comfortably, I am now considering dumping the ESPs entirely and slimming down the software stack on this RPi to that only needed to run the CH and HW control. This topic discusses possible approaches to such an RPi-based control system that meets my needs but once documented could also form the basis of an implementation for others. What I am proposing is to do this in two passes: first more a review and discussion cycle, where input and comment from the likes of @Jeremy Harris, @PeterW, @Ed Davies, @MikeSharp01, @ProDave and anyone else who feels that they can make suggestions and give feedback; the second will be a set of blog posts where I write the entire system up in detail. Switching the power circuits I use Solid State Relays (SSRs) to switch the power circuits for my UFH and HW. All power SSRs pretty much all work the same way, and if you want to know more then Big Clive gives a good tear-down / explanation (here) on YouTube. The CKRD2420 modules that I use can take standard (5V) digital inputs and turn on the power side at the next AC zero-crossing (that is with a delay of up to 10 mSec) if the input goes high. These CKRD2420 modules currently retail at ~ £40 each (compared to the £5 ones that Clive reviewed), but this is the premium for a decent quality approved module from a reputable supplier, and ones that my electrician was willing to sign-of on as part of his 240V installation. The Crydom CKR24 Series Datasheet gives their operation parameters: the input side requires a 4V DC minimum on voltage at up to 12 mA to power the SSR internal opto diode, and power-side TRIACs can drive both resistive loads safely and reliably. This 4V minimum exceeds the general purpose digital outputs of 3.3V based RPis, so some form of voltage / current conditioning is needed to control these SSRs, though the input side of each SSR is already opto-isolated from any transients, it is pretty straightforward to boost each GPIO output to 5V and capable of providing a sustained 15 mA. I prefer to keep my discrete components to a minimum, so a good approach is to use a SN7407N hex buffer driver which can drive up to 6 GPIOs from a single 16-pin DIP package. Note that these buffer outputs are open collector and are driven to 0V when the input GPIO is driven low. The SSR input is connected to the buffer output and 5V, and hence the SSR is on when the GPIO is low. When the input GPIO is high, the buffer output floats and is pulled to 5V using an external 4K7 pull-up resistor on each output, giving a 5V + 5V input to the SSR, that is a 0V difference and switches it off. Note that I previously used an I²C-based port extender instead of a hex buffer driver on my ESP8266 implementation because I had less free GPIOs, however the RPi has a lot of available I/O pins so using these allows me to avoid MCP23008 I²C driver and to simplify the RPi software. One aspect that we do need to careful about is the default state of the relays at power on and before any active software control is initiated. All GPIOs are initiated as inputs, with GPIO 0-8 having pull-ups to 3V3 enabled and GPIO 9-27 pull-downs to 0V enabled. Given that we want any SSRs to default to off at power-on, we need to connect any SN7407N inputs to GPIOs 0-8. The other design issue to note here is that the power TRIACs within the SSR are not 100% efficient; they dump roughly half a percent of the switched power as waste heat, and so will generate roughly 15W of heat per SSR when powering a 3kW circuit. Any enclosure will need to operate within safe temperature tolerances at N × 15W heat output and this will require some form of heat dump design. Collecting Thermometer Readings The digital thermometers are DS18B20s. These are accurate and cheap. You can string as many as need on a one-wire bus which only uses a single GPIO. This bus is fully supported by the RPi's Linux kernel drivers, so the implementation is easy. Hall-Effect Flow Meters These have a 3-wire connector: GND, 5V VCC and 0 / 5V pulse output. Each output needs to be conditioned to be read safely on a RPi GPIO pin but a a 4.7 KΩ + 10 KΩ resistance divider does this job fine. Note is that the Hall output is again open drain. To Follow Physical packaging Circuit designs Software options: native Python vs NodeRED The underlying physics Control regimes and options Anything else that people want to discuss

Unfortunately, it's more a case of US market dominance, and their seemly abhorrence of all things metric. So long as this is a quality set and they fit into your router, then I don't see any major issues. We had our metric router and bit set pinched along with load of other tools in a shed break-in during our construction, so I ended up using an old imperial router and bit set borrowed from a friend who got it as a pass-me down from her dad. It did the job fine, though the main use was for fitting all of the internal door hinges and locks.

It might be worth cabling up the 2 × Willis separately. At the spring and late autumn bookends you might need less than 21 kWh a night and if so then 7 × 3 kWh is a better profile than 3½ × 6 kWh. You could probably get away with a couple of 16A rated programmable timers. If the load is a concern then you could use a couple of 240 V input 240 V × 16A contactors to control the load. The input to these is in mA, so you've got loads of choices of how to switch these, including a wide range of IoT switches.

This is one for your electrician. This is 240V stuff and needs to be Part P compliant. I switch all of my power circuits using Crydom SSRs (sourced from RS Components). IMO you should be wary about most IoT style switches; whilst many are nominally rated at 16A 220V, you will be driving ~12A for many hours per night. These can be driven by 5V TTL outputs, so you can't use RPi GPIOs directly but you can use a level shifter module or something like an The MCP23008 I2C to 8-port GPIO expander which will switch the Crydoms using a 3.3V logic I2C.

Erwin, I gave a couple of very useful links in my Plumbing Design – Part I blog entry, and I found John Heartield's pages very useful. John's widow let me know that he passed away some years ago, but she has kept his pages online in his memory. Pressure drop over a few m of 15mm pipe isn't really going to be an issue, but you want to keep the flow rate in 15mm pipe under 2 m/s otherwise the flow will be audible, especially with those tight 90°s. With both Willis's on you will be putting ~6 kW into a flow rate of say 8 l/min. Do the maths, or now that you've got the real system to test you could just run it and do a subjective check, but you might find it better to do the main loop in 22mm if it is too noisy. BTW, the max O/P temp logged from my Willis in 2½ years running is 37.8°c and the max (out - return) del T has been 4.6 °C. Because perhaps 40% of my flow is diverted through a PHE preheat form my DHW (redundant history) the delta T across the Willis is higher: 7.7 °C. Plugging the 4.6 number into the spec heat formula for water gives an overall flow rate of 0.32 ltr/s through my manifolds which is 0.8 m/s for a 22mm loop (or 2.2 m/s if I'd used a 15mm pipe, and 2 Willis heaters would require 4.4 m/s unless I raised the delta T). This Del T gives a drop of ~ 0.05 °C/m in my pipe at loop flow rate of ~ 0.7 m/s in the UFH pipes. I used the per loop return temps to balance my three loops, Since we run the UFH as a single logic zone, we laid out our three loops to be roughly the same length (all +/- 10% of 95m), so balancing was easy in our case. The maximum difference in return temps between any 2 circuits across a typical 24hr heating cycle is about ¼°c in our case. You've got 9 loops, no doubt of varying lengths and heat output requirements, so balancing in your case is going to be hard, but in the first instance I suggest that you do the same and trim the flow rates so the return temps after 3 hrs heating are roughly the same. As to DS18B20s etc., IMO you don't need lots of these, but they do make understanding and tuning the system a lot easier. I bought a load of the waterproof ones from ebay at ~£1 ea if you buy then in 10s. I just tape mine along the pipe under the sleave logging. As JSH will point these are almost certainly not the pukka Maxim parts but Chinese clones, so you do need to calibrate them. I dropped batches of 10 into an open thermos-style mug with H/W starting at ~40°C and plotted the logged temp against batch average overnight at the water dropped to room temp.. I rejected maybe 1 or 2 per batch if their reading slope was off or the offset was more than 0.4°C IIRC, and also recorded the per device address offset for using in my logging calcs. You can plug these direct into an RPi and use a Python or NodeRED routine to analyse / log them. There are simple worked examples online to tell you how to do this.. No ESPs or Arduinos are needed. I have been thinking about simplifying my current system and dropping my front-end ESPs to run then entire HW / CH system from a single RPi using NodeRED, keeping simple as possible, just in case shit happens and I need one of my kids to do the system maintenance for Jan.

@Pete Have a look at my blog. Some of the photos tell all. Minimal simplicity.

Jeremy, there's absolutely nothing stopping anyone putting the Willis heaters in the manifold loop apart from preconceptions. See my set up. In this configuration you aren't using the manifold TMV for mix down, and I have mine cranked fully open. The only reason that it's there is because it came with the kit and acts as a connector. This isn't the case if you have buffer tank and a heating system that can output more heat that the UFH can sink and stay within design ratings -- such as a 20kW gas boiler, but this isn't the case here.

I've just done a review of our heating costs over the last calendar year. We only use electricity and have SunAmp PVs heated only by E7 for DHW, and a Willis for UFH heating the slab of our 5 bedroom 3 storey house. Based on my Home Automation logs, our annual (Willis) space heating costs for this last year work out at ~£380 p.a. This would fall to around £115 p.a. if we had used an ASHP so maybe an annual saving of around £265. We use an induction hob and 2 × electric oven/microwave for cooking (with a 2 ring Propane for backup and which we've never used in anger) so we have no other need for gas connection. So there would be maybe a 10-year payback for installing an ASHP if I did this myself, plus some extra benefit if we used one with a cooling mode option for the Jul / Aug temperature peaks. We also saved a lot of money avoiding gas installation, and the annual standing charge for gas connection (~£75 p.a. in our area), and the cost of annual maintenance for a gas appliance (something that is mandatory IMO). And note that if we did have an ASHP, then the avoidance of the gas standing charge represents an offset of some 65% of the ASHP running cost. So in our case the gas vs ASHP argument is a no-brainer.

Read my blog posts in modelling this, plus I've also posted my temp logs. The actual behaviour as shown by the logs is pretty much as characterised by the model. On a full night I put about 21 kWh into IIRC 17 tonne of slab, though it's less than that at the moment. The surface gets maybe 3°C warmer. At peak the return flow temp gets up to maybe 32°C after a 7hr heat. Unlike Jeremy's control regime, I don't use the ∆t to limit the heat input. I can't see the need for this when the 2.88 kW output of the Willis is it's own limit on this same regime.

As an addendum to Jan's comment, we have a lot less glass than JSH and so get no material solar gain in the winter months; we also keep our ground floor at around 22-23°C. Hence we have about 3 months where we need lot of heat input and another 3 months where we need some, though my control system does the heat calcs on a daily basis with all (or the bulk) of heating using the 2.88kW Willis during E7 off-peak, and rarely need heat input beyond this. We also use the UFH slab as the primary thermal battery. I don't really understand twhat is gained by using a SA as an additional heat battery in series with the slab.

Ours didn't have the orientation diagram on it and IIRC minimal fitting instructions. The arrows indicated water flow and there was no "this way up" orientation arrow. The gas accumulation issues are completely different for the standard in unvented cylinder installation where the Wills uses thermal circulation in a potable water environment where the water turnover might be ~1m³ / day, but in our usecase we have about 35 ltr of water in our UFH loops -- the same 35 ltr as 2½ years ago when it was first flushed and bled, so there is no further introduced gas to accumulate. I can't see where this ~100 cm³ gas could come. Whilst the inverted position might avoid a potential gas trap, IMO the device can't have been designed for this because of issues of electrical safety: there is no appropriate IPS protection for the wiring contacts and thermal cut out so any seal failure on the element ring in the inverted position would result in electrical shorting. TBH, I think that there are also pragmatic maintenance advantages in having the heater threaded end up, as the element is pretty much at the highest point in the circuit and being upright it could be replaced with minimal water loss if the element failed, thus simplifying refill, flushing, repressurising the system. Incidentally our thermostat cutout is trimmed to ~50°C and I can't recall logging a peak temperature over ~35°C since the system was commissioned. Notwithstanding this, the next time I do some maintenance to the system (possibly over the summer) I might add "cracking the coil" with a heater element key to see if it bleeds any gas to the todo list.

At the moment we are running at about 70:30 E7 to peak rate, but that 30 % has little to do with direct space heating. It's all the other energy use: cooking, lighting, appliances etc., though except for HW going down the plug hole, this generates waste heat that does eventually heat the environment. Because we are currently limited to 2.88 kW input to our UFH, when the average outside temp falls below about 4°C my heating control will top up using peak rate electricity, but this is a tiny % of our total charges. We run our ground floor at ~22.7 ± ½°C ripple. The upstairs floors are a degree or so cooler.