Leaderboard

What would be really useful for people like @joe90, @jamiehamy and me would be to some of those that have been through this such as @jack, @Calvinmiddle, @JSHarris, @PeterW, @PeterStarck just list off their feed stack in terms of components used that is from the MDPE coming in to the split of feeds to DHW and DCW, plus maybe any comments on any retrospective +/-s. And not forgetting the pros here like @Nickfromwales

Like quite a few others here, I bought my MVHR and ducting from BPC. As part of their package they supplied two ventilation boost rocker switches, to allow remote boost activation of the MVHR. Having left in drawstrings for cables, I'm now at the point of deciding whether to bother with these switches at all, as I can program a timed boost to the ventilation rate to coincide with morning showers (this is in addition to the inbuilt humidity sensor), and I will always have the ability to boost ventilation rates from the master controller which will be located centrally in the house. Firstly, just looking for any feedback, good or bad about the switches and whether those who have installed feel they are worthwhile. Secondly, on an installation note, BPC instructions indicate a 1mm 3 core cable from switch back to the MVHR unit. Which terminal connection on the unit did you use to connect these switches?

The best educational H&S video I have seen. Ahem.

I'm on Ebico for the house just now - iirc it's 16p/kwh and no standing charge. It's run and managed by SSE and suits us perfectly for the build. I think we have a tenner DD setup which is far more than we need.

Two comments: Repeating what some once said to me: "when have you ever been f**king humble in expressing your opinions?"; stick to IMO That's what I do. IMO, this debate isn't about real risks, it's about passing the inspection.

Ok, let's clear this up. Situation could be.... Little Timmy decides for whatever reason to put the shower head in the ensuite in the loo. He walks off and plays Minecraft, blissfully unaware that the shower wasn't fully turned off after its last use. Meanwhile.....a plumber turns up and greets mum downstairs. "I'm here to fit your new kitchen sink tap luv". He turns off the stopcock ( which has the non return valve above it ) and opens the kitchen sink taps to let the hot and cold drain down. He goes upstairs to open the taps in the master bathroom and heads back downstairs to fit the new tap ( unaware that there is an ensuite ). After fitting the new tap, the plumber ( good looking welsh guy most probably ) turns all the taps back off and turns the mains on. He tests for leaks ( Da iawn, wrth gwrs ? ) and collects his payment and leaves. Sounds ok .........? No. With the water turned off and the kitchen taps open a vacuum was created. That allowed water to be sucked ( syphoned ) through the partially open shower valve, into the shower hose, out of the toilet, and then into the entire cold water system. It then has time to contaminate the pipework for the time the plumber took to fit the new valve. Timmy then has to brush his teeth upstairs before bed and the water that first comes out of the tap may have been part of yesterday's evening meal. Mum then fills the kettle and that's not great either. Even worse she fills the ice cube tray at the same time and the American fridge pulls it all in too, to be dispensed for the rest of the evening. @ProDave, that tells you how much protection your 3-4 NRV's would provide in this situation. ZERO ! ? Admittedly, the above is absolute worst case scenario, but hopefully allows you all a slightly rough as toast scenario / example to help understand why that type of reg is in place .

That's why it is in the Boffin's sub-forum but when JSH, ST and anyone else with informed views and I have reached a broad agreement to all the geeky shit and agree on conclusions, then I'll do a blog post summarising the "how to" in a blog post or mainstream topic.

Yep. Also we have antislip paint which seems to work OK. Yes expensive, but mine (Cuprinol) came as a Wickes BOGOF which mitigates the £25 per can. For a decking that high you *must* have a pond and a plank, Blackbeard style.

In general no requirement for earth bonding with an all plastic pipe installation, but there are some areas where if you have metal pipe sections and electrical equipment (like a water heater) you still require main bonding. There's a useful paper that Hepworth sent out, and that is now buried on their website about this. It wasn't written by them, but by an IEE chap, but it's pretty much chapter and verse on what should and should not be done. It also gives the reasoning behind the decisions, which is helpful. Plastic pipe earth bonding requirements

Terry that is a brilliant idea, no reason to re invent the wheel so to speak, I personally would be very gratefull for any recommendations as I can "do" plumbing and understand many principles but by no means an expert. For example Terry I have no idea whatsoever what you are talking about in your thread on modelling your slab ?.

I am going to have to reread all that a bit more, and then a bit more again. I had forgotten what was trying to be modelled, which makes a difference to how I will read it.

Get it changed immediately before your rights expire. . When they deliver the new one examine if closely for similar defects etc. If found, at least you'll have a BNIB unit which you could sell / return for refund ( if possible ). Dont fit it until your 100% happy with it. Deal with Stanley direct and explain your concerns, but only after you have the new unit in your possession. My mates hot tub had multiple failures so he fought for a replacement. It had to be craned over the house at his cost. After a load of grief off them he agreed to pay the new crane fees, begrudgingly. When they arrived with the hot tub the crane was set up in the street ready to go. He told the spa Co that he only had 15mns of crane hire left and the original tub was still being disconnected, so they agreed to crane the new one in immediately. As soon as it hit the ground and the crane was detached from the strapping he escorted the spa Co staff out of his home and told them any attempt to reenter would constitute trespass. He retained the original spa and told the company if they wanted it they would have to rearrange the crane at his convenience and that would only be when the new one was proven to be working properly. They never came back and never contacted him again. He sold the u/s one for spares to recover his time and losses. Took 7 of us to 'hedge hop' it over the neighbours gardens. Moral of the story, if they can get off the hook they usually will, so act sooner than later. .

Replies to Qs The GNU gfortran compile is part of the GNU gcc bundle so adding it takes seconds on a Linux (or Cygwin if you run Windows) if you haven't already got it installed. It is Fortran90 and I use the F90 features, so it is a little different from the old F77 that we both used last time in anger. I initially wrote the early versions in C but the C language just gets in the way of reading the code and understanding what is going on. So I moved it over to Fortran as this sort of application fits it so well. So I prefer this. As to the GUI approach, all of the parameters are just that, a block of Fortran parameter declarations (one of the F90 additions and which the compiler treats as compile-time including things like real,parameter::pi = 2*acos(0.0) which is nice). If I want to change anything, just tweak the parameter block, save and recompile; it only takes a couple of seconds. So no fancy input parsing; I leave this to the compiler. As to the output I simply dump CSV data to stdout, and post process this with (LibreOffice) Calc. for example my main output code is: write(fmt2,'(A,I4,A,I4,A)') '(I5,1H,,I5,',2*nRings+2,'(1H,,F8.4),1H,,F12.4)' do i=1, nSections write(unit=*,fmt=fmt2)iter,i,Twater(i),(T(i,j),j=1,nRings),(W(i,j),j=1,nRings),Wair,Q(i) end do (The first line is standard hack because you still can't provide a variable count in a format repeat group.) Sometimes I might use a shell filter such as grep -P '^\s*\d*,\s*50,' /tmp/a.lst > /tmp/a1.lst as a post process. (This picks out all lines with the second field = 50, that is the midpoint in the 100m loop), to keep the Calc file smaller. If such a filter is useful, then it's trivial to add it to the code as an extra line or two of logic. Nick I'll cover this in my extended discussion below. Basis of the model (This section can be skipped by normal mortals, but this in the Boffins sub-forum after all.) My passive slab has ~73m² of concrete 0.1m thick – that is roughly 17½ tonne of concrete in total across the footprint of the house, but the slab also contains another 10 tonne of perimeter beams, cross bracing and steel beneath this which all adds to its overall thermal capacity. Some 55 m² of this is covered by Under Floor Heating (UFH) pipe placed roughly 50 mm below the slab surface in 3 loops, each just under 100m in length and laid in the standard “double back” spirals used in most UFH designs. There is 300mm EPS below and to the sides of the bulk of the slab, so the main (radiant) losses are into the living space above it. Accurately modelling this type of design is practically impossible because of the uncertainties introduced by the steelwork within the slab, and of huge computational cost of executing a time-varying 3D spacial model. So what I wanted to do here was to focus on the time dynamic and heat propagation effects around the pipe run themselves, by modelling an approximation which is more easily simulated but which will have the same macro characteristics of the heat flow through a real concrete slab. So what I am going to focus on in this model is an approximation which is a 100m long homogeneous tube of concrete some 150mm in diameter with a hole running through the middle some 14mm in diameter. Water is circulating through the centre hole at 1 m/s and is heated by a 1kW heater at one end. The surface of the concrete tube is radiating heat into a black body at 20°C at 3.5W/Km². The initial temperature of the concrete is 20°C. My reason for choosing this radially symmetric geometry is a pragmatic one: that this type is it is radially symmetric and therefore computationally solvable with a 2-D approximation, (and its steady state can by analysed analytically). The 3 × 100m pipe runs have a total volume of 5.3m³ which is pretty much the same as the piece of the slab covered by UFH (albeit roughly half of the total mass). The total surface area of the 3 tubes is 141m², which is roughly double the radiating area of the slab, so halving the normal approximation of 7W/Km² will give a similar net heating effect as the actual slab. The 1kW × 3 tube arrangement is directly comparable to the 3kW inline heater that I plan to use initially. So overall this model is good enough to explore some of the issues and performance characteristics that I want to quantify. I approximate the concrete mediums by a set of concentric rings at a uniform mesh interval. As previously mentioned, since the radial temperature gradient is nearly 3 orders greater than the axial gradient in practice I can practically ignore the cell heat flows though the concrete and assume that the only axial heat flow as as a result of the circulating water. Removing these terms also allows me to increase the ∆l (intervals along the pipe) whilst keeping the evaluation stable. By assuming that each material layer is a fixed multiple a fixed thickness ∆xi and by using a fixed time step ∆t, the solution can be approximated on a fixed (∆x, ∆l) mesh using a n∆l × 1-D implementation of a formulation of a simple delta approximation to the 1-D heat equation. This formulation uses 2 material properties that are specific to each material in the wall: VHC the (per) Volume Heat Capacity, which is just a product of the density ρ and the heat capacity cp K the thermal conductivity It also is expressed in terms of the heat flow, W, at each layer boundary and of the temperature at the centre of each layer, so for the jth section of concrete pipe, Tij is taken at midpoint between boundaries i and i+1 and conversely Wij is calculated on the layer boundaries between temperatures Ti-1 j and Tij. Note that a positive W indicates a left→right heat flow (that is T is decreasing in the direction of increasing i). The basic equations internal to a given layer are straightforward and documented in the source code. I've left discussion of the heat equation itself to Wikipedia, etc. Analysis of the Results and What This Tells Me This modelling the slab as three pipes of concrete has some strengths and some serious limitations. The main advantage is that this gives us some understanding how the slab reacts to a prolonged heating period followed by the heat source removal. What it is missing is that this radial approximation will start to break down as the heat diffuses through the slab, and in particular in my slab where I have large internal and perimeter ring beams which add another 70% thermal capacity of the slab and I'll return to this point later. When the 1kW heat is applied it takes a few minutes for the return to start to lift in temp because the slab is cold and sucking all of the excess heat out but over the first 15mins or so it then settles down to a quasi steady state where the 1 kW generates a lift in circulating temp by some 1.6°C so there is a steady thermal gradient of roughly 0.016°/m along the pipe. The radial profiles then slowly rise and the whole temperate slowly increases to support pumping 1Kw/100 = 10W per metre into the concrete. The temperature profile of an individual 1m segment looks very much like the analytic solution to the radially symmetric 1-D heat equation because the second order terms are so small. (I've updated the first graph above to show the temperature profile of the water over time for the start, middle and end of the pipe, and which shows this.) The ambient air temperature plugs into the outer boundary condition. The external heat loss is represented by a radiance of 3.5W/Km² at whatever the external Δt is. This is a bit of botch because the real value is nearer 7 but only half the area is exposed, so this gives the right ballpark for the BC. As soon as the heat is removed, the slab rapidly transitions into a different mode. Since the radial profile is no longer being pumped from the centre, the 1/log(r) type profile starts to relax back to a more uniform profile as the heat start to spread more uniformly through the pipe cross-section. The circulating water now acts to redistribute the heat created by the 1.6° gradient pretty uniformly along the concrete "pipe" and another hour or so the difference along the pipe is minimal. It is really by this point, say 5 hrs, in that the radial approximation begins to lose its validity. The slab is acting far more as a uniform plate some 75mm thick, insulated below and radiating above. In my case the heat energy will also started to get "cached" in the beam underworks partly assisted by the rebar acting as heat pipes. So whilst the general shape of this curve will remain the same, I think that my real slab will be more sluggish in dumping heat so the curve will be flatter. My Conclusions You can use the slab as a buffer If you want to pump 3 kW into the slab then you need to accept that the circulating water temperature will rise of its own accord maybe 5-7°C more than ambient. Alternatively if you limit the Δt a few degrees than the rate at which you an pump heat into the slab will be limited accordingly. Measuring the flow return temperature is the easiest way to instrument the overall temperature of the slab (so llong as you ignore any heating periods and for maybe 3 hours after any heat input.

That's an oak, possibly 35-40 years old from that photo and very rotten ..!!

It was an accident waiting to be! Check the innards: Sometimes things do happen for a reason....if this BC episode wouldn't have happened, my house would be sitting beside a highly dangerous tree!

What is happening at between 22.5 and 23 (I assume °C), the heating side (the left) looks right, but the cooling side (the right) should not have such a pronounced change in it. How does ambient air temperature affect it?

Probably as good a place as any to impress upon folk the need to strip and clean these filters / strainers directly after commissioning, and again after about a month or two of normal service ( use ). Any crap in the incoming supply (new MDPE etc.), or solder, flux, PTFE tape and jointing compounds can get to / and sit on the valve seat and cause it to malfunction. The effect of that failure is for mains pressure to get past the non-sealed seat and damage upstream equipment.

1) Reliance 2) Caleffi 3) Ariston

I just love that you used Fortran! Last time I used it was Fortran 77 running on a DEC Vax, to model 3D weapon trajectories from an instrumentation pack that used three rate gyros and three orthogonal accelerometers, using the Runge-Kutta method to resolve the iterative problems created by multiple double integrations to derive position, when the weapon was rolling, pitching and yawing, so changing the accelerometer references with every sample. Oh what fun that was, when we realised that the model was unbelievably sensitive to the entered initial conditions. It started a whole new project in acquiring very accurate initial condition data. Your graphical results look promising, any chance of the source code? I've been looking for an excuse to install GNUFortran and have a play............................

Twin turbo is what it stands for here.