Part Thirty Four – Things Are Going Slowly – And Some Hot Water Stuff

This thirty fourth entry was published originally by JSHarris on the 29th November 2014 and received 1,256 views on the closed forum


I seem to have been working hard on the internal side of the house for weeks now, but with little outwardly to show for it. At least all the flooring is completed, and the ground floor architrave and skirting is pretty much all done (except my study, which is still a store room), and most of the upstairs architrave is on. I am regretting opting for oiled oak internal joinery in some ways, as it’s taking a LOT longer to fit than plain old painted softwood. Looks OK, though.

We made a late change of plan and decided that, rather than have a fitted wardrobe along one eaves wall of the main bedroom (which would have hampered access to the service space in the eaves, via the small doors I’ve fitted) we’d have a walk-in wardrobe. Space was a bit tight, in terms of depth, because of the location of the bathroom door, but by making a thin partition wall I think I’ve managed to make it work OK. The wardrobe has ended up around 3.6m long by just over 1m wide internally, and this seems to be OK for loads of hanging space, plus shelves and drawers at the ends and enough room to get in sort of sideways at the hanging space. I’ve fitted another oak door and lining in the centre, so the right hand side is a “hers” space and the left hand side is a “his” space, with room access the “crawl door” into the under-eaves service area. Each side has 1.2m of clothes hanging rail, plus around 1.2m² of shelving/drawers, plus storage at floor level for shoes etc. It’s more wardrobe space than we currently use, I think, so should be OK. It’s taken a fair bit longer to think up a way of building it and actually getting the frame in and skinned than I thought it would, but has ended up being a lot more solid than I expected, even though the partition wall is only 52mm thick in total.

Going against advice, I decided to build it as if it were a monocoque structure, to get stiffness and solidity without using thick framing timbers. 90% of the stiffness comes from the glued on skins of 12mm MDF, which are held apart by closely framed thin battens (18 x 28 on edge). By gluing the whole lot up with PU bubble glue I’ve ended up with bonded edge joints between the MDF panels, as well as a very rigid structure, plus it’s heavy enough to give a reassuringly solid feel. Matching the paintwork proved fairly easy, I just primed the MDF with diluted PVA that soaked in well to seal it, used some fine filler to hide the joints and odd ding and then painted it with emulsion. So far I’ve only painted the inside, but you can’t see any difference between the painted skimmed plasterboard and the MDF. I think the secret may be to bond the sheet edges well, with strong adhesive, to prevent any cracks from opening up, plus there are battens under the joints as well. I should be able to add some photos next week, things are still a bit too messy to take photos!

The house still doesn’t really need much, if any, heating, as it seems to get most of the heat it needs from solar and incidental gains. I have had the floor heating on, with the slab being heated to 20.3 deg C by the ASHP and this has been plenty to maintain the house at 20 deg C overnight. During the day it still gets over 21 deg C from just me working plus a bit of occasional solar gain. One disappointment is how much the PV output has dropped with the onset of winter. It’s not really enough to maintain the 260 litre thermal store at 60 deg C, even with no significant hot water use, so is really barely keeping up with the thermal store losses. The preheat system from the 70 litre heating buffer tank works well, and will fairly quickly get the thermal store up to 35 to 36 deg C from the ASHP output, just via the 28mm thermosyphon system. I was going to rely on using the immersion heater to boost the thermal store, but have had second thoughts, as it’s clear that the standing losses are higher than desirable, even in cold’ish weather and the excess PV can be negligible on some days. The standing losses are proportional to the temperature of the water in the thermal store, and are around 1.4 kWh per day when it’s at 70 deg C. If I was to run the thermal store at 45 deg C, then the losses would halve, to around 0.7 kWh/day, but the available water hot water capacity would also drop a lot as well. I intend to add some more insulation around the thermal store to lower the losses, which will help a bit, but the real problem in winter is that it is going to struggle to get hot enough without the immersion being on for fairly long periods at night. I looked at switching to E7, or a similar off-peak tariff, but the added cost of peak rate electricity make it overall al more expensive option. The real issue is the relatively high ratio of standing loss energy versus useful hot water energy.

To get some solid base data on our major hot water requirements, I measured the shower flow rate of our new shower (ran it into a ten litre bucket and timed it with a stop watch) at 9.2 litres/minute, and I also measured the delivered water temperature of the thermostatic mixer shower at our current house (jar constantly filled to overflowing from the shower with a thermometer plonked in it), which delivered a constant 38 deg C. Neither of us ever adjust the temperature setting, so it seems that 38 deg C is our preferred shower temperature. I take around 4 to 6 minutes to shower (depending on whether it’s a hair wash day or not) SWMBO seem to take around 10 minutes per shower. Showering is far and away our biggest hot water demand, the only other hot water use is hand washing plus a very occasional bit of hand dish washing. So, overall, we’ll use around 140 litres per day of water at 38 deg C. The incoming cold water is around 8 deg C at worst (currently it’s around 10 deg C), so our daily hot water energy use, in terms of sensible heat, is about 4.88 kWh. With the thermal store at 70 deg C it would rise to around 6.3 kWh because of the standing losses from the store. In simple terms we’d be “wasting” around 23% or so of the energy used to heat water to 70 deg C in the thermal store as losses (not true waste, as it does warm the house, but that heat isn’t either very useful or in the right place, as it’s next to a bedroom).

I can cut the “waste” heat down to about 12% by running the thermal store at 45 deg C, but then I can only produce a small volume of water at 40 deg C before the output temperature starts to drop off, because of the reduced store of sensible heat (around 25 litres or so of hot water at 40 deg C). However, I have a bigger effective reserve of water in the thermal store at 35 deg C, because of the 70 litre buffer tank that’s preheating it. What I needed was a way to be able to allow the thermal store to deliver hot water at a mixed down 40 deg C when there was enough spare PV to heat it right up, but also have an economic* boost system that would add enough heat to the water when the thermal store thermostatic mixer valve (TMV) allowed water that was cooler than the set 40 deg C to flow.

* Economic in terms of the balance between capital expenditure, life and running cost

After a bit of digging around I found a couple of instantaneous electric water heaters (both German made) that use a different control system than that used by 99% of electric showers. Instead of controlling the water flow rate to adjust the temperature, they electronically control the power to the heating element. The big advantage this gives is that they will provide exactly the right boost needed to heat the hot water to the set temperature, even if fed with water that is pre-heated. Best of all, these heaters are smart enough to only turn on if the water actually needs heating, so if they receive water at 40 deg C, and are set to deliver water at 40 deg C, they do nothing, but allow all the water through (at flow rates of up to 25 litres/minute). Adding one of these to the TMV output from the thermal store allows even low grade heat from the ASHP to be efficiently boosted to 40 deg C.

At first thought, using direct electricity to heat water seems daft, because of the relatively high cost of electricity. However, there are several things that work in its favour. Firstly, is very, very close to being 100% efficient, so you only use the exact amount of electrical energy needed, when you need it, with very little waste. Secondly, the capital cost of direct electric heating is relatively low. Thirdly, direct electric heaters tend to have a fairly long life, as they are pretty simple things, with only one moving part (a flow switch). Finally, they are small, around the same size as an electric shower, so can be fitted pretty much anywhere.

To illustrate the effect of one of these instantaneous heaters, the graph below shows how the thermal store hot water output falls off with decreasing temperature within the store (there’s around 2 deg C temperature differential between the water in the store and the water available to the TMV). This graph show the performance of the thermal store on it’s own, the performance with a Stiebel Eltron DHC-E 8/10 set to 7.2kW and the performance with the same heater set to 9.6kW (it has two maximum power settings, set using an internal link on the controller):

The big advantage with the boost heater is that it allows the hot water system to deliver hot water at 40 deg C right down to a thermal store temperature of 29 deg C. This gives an extra 3.92kWh of usable sensible heat, excluding the buffer store contribution, or about 4.4 kWh of extra sensible heat allowing for the buffer store. This also means that we can get enough daily hot water from just the ASHP heating the buffer store and thermal store to around 35 deg C, without using the immersion at all, as the recovery time between showers from the ASHP alone will be enough to provide warm enough water for the second shower (with an up to 9.6kW boost). Luckily SWMBO showers an hour or two before me in the mornings, and my showers place the least demand on the system, so this should work well. The ASHP can heat the thermal store and buffer tank, from the lowest usable temperature of 29 deg C to 35 deg C in about half an hour when running at full output.

I’m still planning on adding extra insulation to the thermal store, essentially a 60mm thick PIR foam box built around it, with a cut out to prevent the immersion lid getting too hot, and with taped joints so that I can open it up if access is needed at any time. I’ll keep the immersion night time boost timer and thermostat, but set that to cut in when the centre of the thermal store drops to about 30 deg C, as an emergency back up. When the sun shines the thermal store can still heat up to 65 deg C from excess PV, and work normally as the primary hot water system.

In terms of running cost, the ASHP will preheat with a high COP; at least 3 with a 35 deg C flow, closer to 4 much of the time. This means that the 260 litre store and 70 litre buffer can be heated from the lowest temperature it’ll get to of around 29 deg C to 35 deg C for around 11p, and then the boost heater will cost around another 20p for a 10 minute shower. Overall this seems an economic way of providing a winter hot water system, with minimal capital outlay (the boost heater was around £240). It’s cheaper than opting for a higher flow temperature ASHP, which would almost certainly have a low COP plus an inline boost heater anyway (our inverter driven monobloc doesn’t have a boost heater, but many do to get the higher water temperature needed to feed a hot water tank at 50 deg C or so without excessive icing up during the critical outside temperature range of 0 to -4 deg C).

There is some merit in just fitting instant electric heaters at all the outlets instead of running hot water pipes around the house, too, especially for all the low volume hot water demand outlets (basins, kitchen and utility room sinks). The reduced standing losses, availability of instant hot water and simplified plumbing are good enough reasons to make such a system worth considering. For the bath/shower demand I think I’d opt for an unvented indirect hot water tank, with the coil fed from the ASHP buffer and an immersion to boost it up from excess PV when the sun shines. I’d feed this to an instant water heater such as the one I’ve just bought, to give plenty of hot water to the bath or shower even if there’s been no excess PV to use. The advantage of an unvented hot water cylinder is that it can be a bit smaller and you can use more of the stored hot water as a preheated source for the instant water heater than you can with a thermal store. The downside is the need for a G3 sign off, which means getting a plumber with the right ticket to install it.

declan52 29 Nov 2014 08:35 PM:

 Did you use the mitre glue to build the architrave up in one piece.


 jsharris 29 Nov 2014 09:44 PM:

 “declan52, on 29 November 2014 – 08:35 PM:, said:

Did you use the mitre glue to build the architrave up in one piece.”

 I did use the mitre glue, and found it to be great stuff, but I didn’t end up using it as intended. I tried making up the architrave flat on the floor, using mitre bond to fix the mitres, and it works OK. The snag I found was that I didn’t want to use any fixings at all on the architrave, as I finished them with oil, and I found it easier (at least on the hinge side) to fix the verticals in place first, then fit the top.

 Because it was impossible (or at least damned difficult) to cramp the verticals on the hinge side, I ended up using a two adhesive technique on the hinge sides, that worked well (but is probably frowned upon by the experts). I’d first run a bead of the solvent free Screwfix grab adhesive down the wall/edge of lining, then I’d run a bead of mitre bond only down the lining. I’d then put the primer on the architrave and quickly line it up and press it into place (a toe, knee and hand job). Holding it for 20 seconds or so was enough to get the mitre bond to go off, holding the architrave in place whilst the grab adhesive went off.

 The verticals on the other side could easily be cramped in place, so for those I just used grab adhesive. The horizontals were fitted with grab adhesive on the back and mitre bond and primer on the mitres, so they could just be plonked on, held tight for 20 seconds and then left for the grab adhesive to go off.

 The results are OK, although you can see the difference between the first ones I fitted and the last ones, after I’d got my eye in and perfected the technique. It’d have been a lot quicker to pin them all in place, but I was adamant that I wasn’t going to have a load of pin holes to fill and stain to match.


 declan52 29 Nov 2014 10:04 PM:

 Its a technique more aligned to using nails. You must have some patience to do all that just for door architrave. I think after doing the first one I would have given up and used nails.


 jsharris 29 Nov 2014 10:18 PM:

 It depends how much you dislike the look of filled pin holes in plain finished oak!

 Mind you, it may well be why doing all this detail stuff is taking me so long………………….


 NSS 29 Nov 2014 11:46 PM:

 Hi Jeremy, apologies if you’ve covered this previously, but I assume you have no gas boiler or other means of providing DHW other than the ASHP & immersion heater? I’m interested because I’m also going to be relying on this combination so the direct heater you speak of may be something I should also consider adding.


 jsharris 30 Nov 2014 10:11 AM :

 You’re right, our only energy source is electricity. We didn’t even have space for an LPG tank or oil tank (not that I really wanted either) so knew from the start that the system has to be an all-electric one.

 The original idea was to use excess PV alone to heat the thermal store for hot water in the summer (something it will do with ease, based on the performance of the PV system this past summer) and hope that the combination of pre-heating the thermal store to 35 deg C from the ASHP, plus excess PV, would be enough to provide most of the hot water in winter. The problem is that it isn’t going to be anywhere near enough to provide the winter hot water, so I needed an alternative.

 Another problem is that I opted to use a combination thermal store because I didn’t want to get involved with having to use a G3 qualified plumber to install an unvented hot water cylinder. With hindsight this was a bit shortsighted, as the thermal store is not effective when run at low temperatures. I may yet end up removing it and fitting an indirect unvented cylinder, with the coil driven from the buffer tank as preheat and an immersion fitted in the bottom for the excess PV heat input.

 The snag with the thermal store is that it doesn’t really let you use all the warm water in it as effectively as a unvented cylinder, as it relies on there being sufficient temperature differential to run the heat exchanger coil inside the store that delivers the hot water. If I had a 260 litre unvented cylinder heated to 40 deg C, then I could draw off a fair bit of that 260 litres from it at that temperature, with the replacement cold water refilling it from the bottom. With a thermal store at 40 deg C I can only draw off a few litres before the hot water temperature starts to drop below 40 deg C, because the incoming cold water to the coil starts to drop the whole thermal store temperature as soon as a hot tap is run.

 The thermal store running at a low temperature is virtually free from any legionella problem, though, because of the small volume of water in the heat exchange coil that gets flushed out and, because it’s usually corrugated copper, is a more effective biocide than a tank. An unvented cylinder run at a low temperature would need an anti legionella system, perhaps a timer that turns the immersion heater on to get the cylinder up to 65 deg or so once every couple of weeks to kill off any bugs.

 What I have found is that the ASHP is pretty good at preheating the hot water system, and seems to be running pretty efficiently when doing this (at a guess I’d say that it’s been running at a COP of around 4). I did try running the ASHP in hot water mode (which increases the flow temperature to 50 deg C) but it was very obvious that it was working a great deal harder and the power input increased dramatically more than I would have expected for such a modest temperature increase, which makes me suspect that the COP had dropped a lot.

 If doing this again, knowing what I’ve learned so far, then I think I’d go for one of the following options:

 A system rather like that I have now, but with an unvented indirect cylinder for hot water. I might even have fitted that downstairs and used it as the buffer for the heat pump, rather than have the two tank system I have now. I’d stick with the immersion and excess PV diverter for summer hot water and add an instant water heater to boost the hot water temperature when needed. I’d also have an anti-legionella timer to run the immersion once every couple of weeks to kill the bugs. I’d still use the radial plumbing system I have for hot and cold water to each room that needs it, run from a manifold.

 Basically the same system as above, but with a smaller unvented cylinder (just enough to supply about 1 1/2 showers, allowing for recharge between) and no hot water feed to the kitchen, utility room, WC or bathroom basin taps. Instead I’d use something like this: http://www.stiebel-e…s-water-heater/ at each of the low volume hot water outlets to provide the small quantities of hot water needed instantly, with virtually no losses. I’d stick with the bigger DHC-E 8/10 to run the bath and shower hot water feeds. The advantages of the under-sink/basin units is that there is only a cold feed to each room, hot water is available instantly on demand and there are virtually no standing losses. The reduction in standing losses is probably enough to make up for the use of full price electricity. I did some rough and ready testing on hand washing, and found that our washbasin and sink hot taps take around 30 seconds to 1 minute to run hot. The average length of time the hot water is left running is around 30 seconds or so at the most for hand washing. The conclusion I drew was that we were wasting more hot water in standing pipe losses at these outlets than we were using most of the time.


 NSS 30 Nov 2014 10:21 AM :

 Thanks Jeremy, that’s really helpful. Am particularly liking the second option.


 TerryE 30 Nov 2014 10:21 AM :

 Thanks for this roll-up of the discussions in the various related threads. Extremely useful to those interested. Your monocoque assembly technique for the walk-in is an improvement on a similar one we used for our existing one and we’ve had absolutely no problems with it in the 10+ years that it’s been up. I’ll use your method but with Fermacell instead of MDF.

 I will look at the figures, but the one area that we might vary from you on is the economics of E7. We use it at the moment and it pays overall. With a little planning, you can move an awful lot of your demand into the cheap rate: DHW, washing machine, dishwasher, … And if we go the PV route then a lot of our daytime consumption will free anyway.

 Have you got a good handle on the base running costs for the UFH circulation pump and the MVHR?


 jsharris 30 Nov 2014 11:10 AM :

 The UFH pump runs mainly during daylight at the moment. It’s on a timer so it comes on at 06:30 and (at the moment) goes off at 18:30, so around two thirds of the time it’s on, even on a dull day, it’s running from the PV system. The pump is running at it’s lowest speed (it’s a Grundfos UPS2: http://uk.grundfos.c…oduct/ups2.html) and is probably drawing around 20 to 25 Watts.

 The MVHR similarly runs at it’s lowest fan speed all the time, but is running 24/7. The spec suggests that the power consumption is around 25 to 30 Watts at this speed, perhaps a bit less.

 So, at our present electricity rate of about 14p/unit, the daily running cost of the UFH pump and MVHR (ignoring any PV) would be around 14p per day, but in reality, even on a dull day, around 1/3rd of that would come from the PV, so the real cost is probably around 9p to 10p per day, less in summer when the PV is producing for longer periods.

 The problem I found when I looked at E7 was that it forced me to use a thermal store, or high capacity unvented cylinder, running at a high temperature. If it wasn’t for the problems I’m getting with the services space being up around 30 deg C from the thermal store heat losses then I may have considered it, but it would pretty much rule out using instant water heating, as the higher day tariff would start to erode away the advantage of having virtually no standing losses.

 We did live in a place in Cornwall years ago that had E7, and did find that it wasn’t too bad if we shifted loads to run overnight. However, with a big PV array it makes sense to run those loads during the day, at least for two thirds of the year. This then means that E7 is really only useful for winter hot water and running the washing machine, dishwasher etc at night, and so any saving is somewhat eroded.


 bitpipe 01 Dec 2014 10:05 AM :

 Very insightful as ever Jeremy. We’ll have a much larger DHW requirement for showering (potentially 3 showers more or less simultaneously) but I completely agree that when it comes to hand wash basins, more time is typically spent waiting for the hot water than actually using it.

 I recall a discussion about using a pumped loop for DHW to ensure that it was always available ‘on demand’, question is whether the losses of circulating the DHW (and the energy used by the pump) would be more or less than the standing pipe losses.

 Are your DHW pipe runs insulated?


 Crofter 01 Dec 2014 02:39 PM:

 I’ve looked into those Steibel Eltron units myself, with a view to bypassing my existing DHW cylinder. However it seems that they don’t like having a thermostat downstream- I guess funny things happen if the flow rate is being choked off after the heater. So I’m just going to fit an electric shower instead.

 In the new house, everything will be on instantaneous heating, for simplicity, eliminating standing losses, and low initial cost.


 ProDave 01 Dec 2014 03:49 PM:

 Following on from the fixings of architraves etc.

 When our present house was built, we paid a local building firm to build the basic shell then did the insides ourselves. If that builder had done the internal fit, they told us the doors would have come already fitted to the frames and with the architrave already fitted to one side. Just slot it in the hole and fit the architrave on the other side, job done.

 I am still debating exactly how to heat our hot water and am still erring towards economy 10 to run the heat pump to heat the hot water tank over night and warm up the UFH in the morning. Then again time the heat pump to top up both again during the mid day cheap rate time. Then using it manually at other times of day. but I would only heat the tank to “just hot enough” to allow headroom for surplus solar PV to heat it further.

 Our hot water tank will be in a cupboard from our bedroom so there’s the same risk of overheating it. I might see if we can configure it so the tank cupboard opens from the en-suite. Less likely to be any complaints about a hot en-suite than a hot bedroom.

 I have covered this option by including an extra core in the SWA from the meter box to the house so we are ready for an E10 supply should we choose that option.

 SWMBO is adamant she wants under floor heating in the en-suite. I don’t want heating upstairs (in the belief we won’t need it) so again I’m thinking of electric UFH in the en-suite to warm the floor in the morning at the cheap rate ready for the morning shower.


 jsharris 01 Dec 2014 07:27 PM:

 Our DHW pipe runs are semi-insulated, in that I fitted them through the insulation in the floor space for much of their length, so they are insulated for more than half their length. This doesn’t make a lot of difference, because even the best insulation will only keep the water in the pipe warm for an hour or so, so 90% of the time you’re effectively heating the pipe run from cold (or at least from room temperature).

 I specifically asked the question of Stiebel Eltron about feeding the DHW supply from this unit to a thermostatic valve. The answer was that the DHC-E units can cope with this without any problems at all, but the hydraulic control version cannot. The answer is to make sure you use the DHC-E unit (or the equivalent from someone like Clage) that uses electronic power control, as this allows use with potentially restrictive outlets.

 I’ve been working upstairs for the past week, and can say for sure that no upstairs heating is needed. I’ve been sweating buckets working in the bedroom on the fitted wardrobe, with no upstairs heating at all and the thermal store turned off to stop heating from it to the upper floor. I’m not bothering with any upstairs heating, but have fitted 150W towel rails in the bathrooms. I’ve tried these, and you really don’t want them on for more than about an hour in the morning, as an hour of 150W in a 3m x 2m bathroom is far more than needed for comfort.



Leave a Reply

Your email address will not be published. Required fields are marked *