Working out a scheme for DHW

[dnb]

I was, until I joined the forum here, fairly set on solar thermal plus backup electrical heating for my new house DHW. Now I am less sure having read some well written arguments from some forum members. So now I need to revise my models and find what is the most suitable technology.

 

First of all, it's a reasonably large house with 2 ensuites, a family bathroom and a wet room down stairs. I have a near teenage daughter and a wife who both adore water. (I know there is a theoretical limit of 125l/person/day but let's suggest, for argument sake that I have several children and several wives while not admitting to any wrongdoing whatsoever...) I also have elderly parents who like visiting their grand daugher(s!) on the Isle of Wight, and may one day forget to go home.

 

The site has mains water and mains electricity. That's it for services. It also has lots of trees, but as stated in other threads, burning things in houses is a 20th century idea. Oil should be saved for use as petrol in classic cars in my opinion.

 

My airconditioning combined with MVHR thread is going pretty well, so let's make the further (perhaps rash) assumption that this will work. The DHW system will therefore be broadly separate from the space heating.

 

The first picture shows the distribution of energy I expect to need each year for hot water. It is based on simulating water heating loads randomly over a morning and evening. For instance, some mornings there are 2 showers and nothing much in the evening. Other times there is one shower in the morning and 2 showers and a bath in the evening. Duration was modelled as an almost Gaussian distribution centred on 8 minutes for the shower, modified with a long tail to the longer duration to model teenage daughers... So on average I am going to need to find 4,800kWh over a year, preferably mostly out of thin air. Does this seem under done to anyone?

 

Thankfully, I have 40m^s or so (let's be a bit conservative) of south-ish facing roof with a 43.7 degree pitch. (Would have been 45 degrees, but this simple change saved me the best part of £12k in the house structure)

 

So, how do I supply the energy for DHW? I should point out that if at any point my wife decides to make use of hot water and there isn't any then I will be fed to the sharks. Nothing like a bit of pressure is there?

 

As I said, my first thought was solar thermal. It makes grand claims of 80% efficiency etc but these are in optimal conditions, not when feeding a thermal store up to high temperatures. But let's assume I fit 9m^2 of themal evacuated tube "panels" with a 500 litre thermal store and backup electric immersion heater, and the system "works perfectly". (Waste heat will be dumped to a radiator on the north roof without incident - it's a perfect, albeit imaginary, world.) Insolation data comes from several years of daily measurements of intensity and duration from a reputable source. This is run on a daily basis against my random usage model for thousands of years to build a decent statistical sample set. One year of this is presented in the second chart showing  how the system responds on a daily basis. This tells me I need to get a decent control system for it all - or write one if all else fails.

 

In the winter, early spring and late autumn, the system needs to use the immersion heater before 6AM in order to ensure I don't get fed to the sharks when the wife wants a shower before going to work - there being no useful sun until 9 or 10AM. This of course means that some days produce waste heat, even in winter. (Telling the wife to plan showers based on the weather involves things that are worse than sharks.) Similarly, we need to add heat with the immersion heater if the day's solar capture is less than we need for the evening baths and showers. The chart therefore shows that much of the time an excess of hotness is generated in order to minimise grid electricity consumption, and that this is a law of diminishing returns because there's no use for heat you can't store and having roof area that is only usefully doing something for 1/10th of the year is not cost effective.  Perhaps a PV only system would do better? It may now be cheaper and certainly there would be less waste especially when a battery bank is considered, although currently these don't seem cost effective - but that is another story.

 

Note that the above analysis might just as well be performed on a PV system. I will do so next time I play with Matlab because everything has changed price since 2017 when I first looked at this.

 

So will 9m^2 of solar thermal and 18 300W PV panels do better than simply putting in 24 PV panels? (The Myford is begging for 3 phase, so I need an excuse.) It would appear my model of solar thermal needs to be updated based on the contents of the forum and some links. I wonder if it would be sufficient to use an average efficiency to derate the thermal panels in the model to get something going quickly.

 

NB: The wife isn't the shark queen and doesn't really want me out of the way (at least until the completion certificate for the house arrives) but she does suffer endlessly for my sense of humour.

 

 

[Onoff]

6 hours ago, dnb said:

(Would have been 45 degrees, but this simple change saved me the best part of £12k in the house structure)

 

Can I ask how that saving comes about please, given such a small pitch change?

[Onoff]

Great thread start btw. As storage seems to be a key I think I'd be using (to avoiding becoming shark bait), that £12K roof saving and working a Sunamp into the mix somewhere. This assuming they've sorted their issues out. They've a longer decrement delay than a TS.

 

I presume you're considering in roof pv to further save on the tiling?

[SteamyTea]

Ah, Matlab, I still feel the cold sweat of university.

ST has a few maintenance issues that need to be considered, PV is, generally speaking, either working or not.

Also it is much easier to route cables than pipes.

You can use a heat pump to raise the temperature of your mains water a few degrees, then for the final push from say 40°C to 55°C use the immersion heater from the PV.

As for the best way of doing this, it is difficult to decide.  Two cylinders is probably the answer.  A lower temp one feeding the higher temp one.

As an example, I have a simple, vented, Economy 7 cylinder with a capacity of 200lt.  I run it at 50°C.  This is generally enough to for 2 showers and a bath per day.

I can raise the temperature up to 65°C when I have people to stay.  That gives me more capacity without having to turn the secondary 'top' element on.

I had a lodger once that was under the assumption that the time to finish a shower was when the water ran cold, left me with a £2000 water bill.

I trained her up to take much shorter showers.  It worked, she could wash in 3 minutes in the end.

Limit the flow of the showers to no more than 9lt/minute.  Mine is 11lt/minute and very nice.

[ProDave]

Okay my thoughts.

 

Forget a thermal store. That forces you to store the water hotter than the temperature you want from the tap, so harder for solar thermal to work (if that is your chosen route) and higher standing losses.

 

Instead use an unvented cylinder.  the water temperature in the tank is how hot it comes out of the tank.  You only need to store water and say 47 degrees, I actually have mine at 48 degrees and that is hotter than I can hold my hand in for more than a few seconds, so why would you want it hotter?

 

I chose a 300L unvented HW tank. That might sound large, but given you are storing it at "usage temperature" so it's not getting diluted with cold very much, I would say that is only just enough when the "ladies" decide to shower in sequence.

 

How to heat it.  Well you can still fit solar thermal, but that only really works under certain conditions and as mentioned has maintenance issues.

 

So how about an air source heat pump. that is what I have, which does heating as well. That is quite happy heating water to 48 degrees.

 

I also have 4Kw of solar PV. That has a (home made) automatic diverter that sends any surplus electricity (that would otherwise be exported) to the immersion heater.

 

I have the heat pump timer set to come on at 10AM by which time there will be at least some useful solar PV generation.

 

The last part of the jigsaw is how to cope with unexpected high demand. What if all the hot water has been used up and you hear the shout "my shower has gone cold".  Well in line with the output of the HW tank is a Stiebel Eltron  modulating 10Kw electric water heater.  Most of the time the water passes through this without being heated, but if the water temperature starts to fall short of your target temperature, then this heater will heat the water real time and save those embarasing "shower gone cold" moments.

[joe90]

I’m with @ProDave as we have very similar setups and mine works perfectly. I have an in-line heater ready to instal if required but it’s not so far. I have a 4Kw ASHP and that does underfloor heating and DHW with a COP of about 4.( a 280 sq m house).  When my system settles down (not long been completed) I am also considering E7 (I have no P.V..) to bring my bills down even further.

[SteamyTea]

Just had a thought while enjoying a hot bath.

Heat loss from an off the shelf cylinder can be quite high.  I calculated that my losses where at least equivalent to my usage, and sometimes higher.

To get around this I added extra insulation.  This lowered the losses considerably, as did lowering the storage temperature.

Worth considering the size of the cupboard that the cylinder is in, and how easy it would be to add extra insulation.

Not to be outdone on the charts, here is May's mean energy usage by hour.

I am usually up by 6AM, so water heating is between 1AM and then.  I have limited the E7 window so that it now starts at 4AM.

 

[Jeremy Harris]

I went through a lot of detailed thermal modelling BEFORE installing a thermal store, using the data available from the manufacturer.  It was an unmitigated disaster, as the manufacturers data was from a standard test method that I was unaware of, and which bore no resemblance to the way a thermal store (or hot water tank for that matter) would be used in practice.  It seems that the test method used assumes that a boiler will be fired up to fill the tank shortly before hot water is needed, then hot water will be drawn off, leaving the tank relatively cool for a long time, until the boiler is fired up again.  This doesn't match the way hot water storage is often used in a house where the water is heated from excess PV generation at all well.  The main consequence is that the real world heat losses are massively greater than the manufacturer's data would lead you to believe.  In our case, the losses were enough to push the temperature in our services room up to over 40°C, hot enough to cause damage to the door into the room.

 

The first attempt at a solution was to add an additional layer of 50mm of PIR foam, as a well-sealed, foamed in place, extra jacket around the tank.  This helped, and reduced the leat loss by about 1 kWh/day, but the loss was still way too high.

 

The fix was to remove the thermal store and replace it with a Sunamp.  The Sunamp has a really low standing loss, so low it's pretty negligible, and this both removed the overheating problem in the services room and significantly improved the overall DHW system efficiency.  Our system uses excess PV generation as the primary means of heating the Sunamp, with a backup E7 boost from a time switch to ensure the Sunamp is charged first thing in the morning.  This latter arrangement is only really needed in winter, when there's very little useful excess generation from our 6.25 kWp PV system, but it's a handy way of ensuring that the Sunamp always has enough charge for morning showers, even if it's been used to heat water for a bath the evening before.

[SteamyTea]

7 minutes ago, JSHarris said:

The first attempt at a solution was to add an additional layer of 50mm of PIR foam, as a well-sealed, foamed in place, extra jacket around the tank.  This helped, and reduced the leat loss by about 1 kWh/day, but the loss was still way too high.

What I did, but added about 150mm.

Part of the reason I limited my E7 window to the last 3 hours of the period was because I use most of my DHW first thing in the morning.

Since changing my cylinder a year ago, I have not got around to fitting any temperatures probes, should have done it at the time as I don't fancy the doing the job now.

One thing I have thought of doing is to fit some probes to the header tank, then I can see how much the temperature varies up there.  My summer DHW usage is significantly higher than my winter usage, so would be interesting to see if it actually costs more.

[joe90]

However a few of us have used tempest pre insulated unvented cylinders and not found the losses very high, my DHW is set to about 48’ like @ProDave and my wife mentioned the airing cupboard is not warm enough. In fact I just opened the door and I cannot tell the difference in temp inside. I did measure the temp in the cupboard fir a while and it was a Couple of degrees warmer than the house.

[ProDave]

Yes that is another vote in favour of an UVC rather than a thermal store, that at least the Telford stainless steel tanks do seem to be pretty well insulated, and if you take care to lag all the connecting pipework it does not seem to leak very much heat into the room.

 

As mentioned the Sun Amp is another candidate. More expensive but very much lower standing heat losses, but a different beast to learn about and how to heat it. I will let the experts discuss that one.

[scottishjohn]

just to add another dimension and possible way to do things

I ran  solar thermal evac tubes on my house for years and I found them very good for dhw  and indeed also helped ufh 

they were removed when i went ASHP-cos i would not get same RHI if they were fitted 

nothing to stop you fitting solar after doing RHI claim ?

anyway what i had was unvented 300l tank which had 

solar coil at bottom ,then UFH coil in middle of tank 

dhw at top --with cold mixing on out let --so i did not waste the heat i stored by adding a lot of cold at taps 

and the parent volume of the tank was heated heated if needed by a lpg boiler 

In truth it was not a large enough tank  cos i could boil it in the summer just with2 x20  solar panels -

 

I fully understand the heat loss argument 

 

I ended up surrounding my tank with 50mm pir pipe insulation segments --stock item as tank was 500mm

this made a major difference to heat loss --but made summer boiling worse .LOL

 

so only real downside was tank too small  

 

80%of the year i got all dhw from solar +some UFH --which due to old house design i needed more than you would with modern house --overheating of house was and never is a problem 

 

 

If doing it again i would have at least 500/800litre tank with same coil configuration and as tall as you can fit --taller+ larger  if you can accommodate it and maybe UFH coil closer to bottom 

I not looked at it yet but maybe ashp coil in tank at that level as well rather than heating all tank--that will be investigated once i have the plot sorted +house design

If you have PV then maybe elec element close to top of tank to boost dhw when solar cannot _+ If you have free energy --it will fill the tank downwards and allow solar coil at bottom to work at low temp in winter when you maybe only get 30-40c on bad days -which  with passiv type house etc is plenty for UFH,as most of you are saying your floor is running at 20-22c

+ 150mm of pir around it and all pipes lagged -above 150mm the extra cost /insulation is not cost effective-so Kingspan man calculated

taller means more thermal stratification to get best solar gain and DHW production  and you do not want the water stirred up -

solar coil would have a "top hat" over it with a pipe going to top of cylinder so hot water goes to top ,if hot enough first ,with side outlets at different levels so it heats where it can 

I know this tank sounds expensive --but balance that against cost of  multiple sunamp units 

Yes PV can be used for other things ,but if you have space for both ,then your PV can do  things other than water heating  , 

solar thermal can do DHW-and  some of the UFH 

I believe there is a good reason for both in a balanced system ,if space available for panels -

I do not think that now PV is here that solar thermal is obsolete-it  has a place along with PV

what i have yet to find is a direct comparison of pv /verus solar --purely for heating water -on A DAY WHEN EITHER CAN DO IT

I suspect solar thermal will be be better -as you make voltage ,transform +invert it before you send it to an element -got to be losses !!

solar thermal is direct+ much simpler 

 I think the previous posts over estimate the maintainence problems 

i ran mine for close to 20 years with no problems  other than those caused  by me  in a round about way 

 

the boiling problems --due to too many panels and not large enough tank -fied  a sensor  which then allowed it to keep running -=-result all foam insulation on copper pipes melted off !!--so that was in excess of 170c

another thermal sensor in another postion as a safety measure would have stopped that -which i then fitted

apart from that I spent nothing on system 

 when i removed it i found some of the tubes were blown --but i am putting that down to my major overheating issue  --but even with 6 out of 40 tubes gone it was still working ok 

the cost to repair would be less than an invertor replacement .

 

 

 

 

 

 

 

 

 

Edited June 11, 2019 by scottishjohn

[newhome]

48 minutes ago, scottishjohn said:

nothing to stop you fitting solar after doing RHI claim ?

 

With RHI you sign to say that the system won’t be altered and it has to be open to inspection (in theory). You can use ST for DHW but not space heating with RHI unless you have a meter for performance on an ASHP from memory. Bloody stupid if you ask me but those are the rules. 

[dnb]

I wasn't planning g on any RHI applications unless it was easy. Letting a short term government scheme drive system design is too much like my day job. I will read the posts in detail tonight when I am not at work. Thanks all!

[scottishjohn]

3 hours ago, newhome said:

 

With RHI you sign to say that the system won’t be altered and it has to be open to inspection (in theory). You can use ST for DHW but not space heating with RHI unless you have a meter for performance on an ASHP from memory. Bloody stupid if you ask me but those are the rules. 

yes you can but ti drops the payment you get for the ASHP --cos rate for solar is  a lot less than ASHP

chances of some one coming is virtually zero -

[newhome]

1 minute ago, scottishjohn said:

chances of some one coming is virtually zero 

 

Yes I imagined that would be the case. Must make an effort to try (again!) to get someone to fix my ST before the summer. It would probably do a fair chunk of my DHW requirement if it worked. 

[scottishjohn]

1 minute ago, newhome said:

 

Yes I imagined that would be the case. Must make an effort to try (again!) to get someone to fix my ST before the summer. It would probably do a fair chunk of my DHW requirement if it worked. 

it will do it all !!

whats the problem with it ?

 

[newhome]

1 minute ago, scottishjohn said:

it will do it all !!

whats the problem with it ?

 

 

Never used to heat the water up quite hot enough for me when it did work previously so it always needed a bit of a top up from electric (hubby found the temperature ok to shower with but I like it skin removing hot lol). 

 

Thought to need the AAV changed on the roof. As soon as you mention roof no one wants to know and I struggle to get trades out here at the best of times. I have the bits (new AAV and new pump just in case) bought from the original supplier of the kit but they’ve been sitting here for a year now. 

[scottishjohn]

evac tubes or panels /

how many --what size cylinder ?

what controller 

can,t be that hard to borrow a roof ladder or make one out of an normal ladder for that job 

or just tie a rope on it and throw it over the roof and secure on other side -

or rent a cherry picker?

I saw a danish one where he mounted panels on the wall vertically --this cut down summer heat a bit to stop boiling --but raised up the winter output +he could get at them very easy 

certainly will be considering that or  a ground array  on next house --getting too old to go on roofs 

I don,t bounce like i used too

last one was on flat roof on garage on frames and set at 54 degrees same as our latitude so fairly vertical then 

 

[newhome]

Evac tubes. 

ECO – 58-1800-30 x 2 so 60 tubes

475l TS

Deltasol BS/4 controller. 

 

I can’t go on the roof myself and no one else round here wants to so stalemate. 

 

[dnb]

10 hours ago, Onoff said:

 

Can I ask how that saving comes about please, given such a small pitch change?

 

You can. The roof span at 45 deg pitch was just over the maximum span for SIPs panels so 3 14 metre glulam purlins were needed to support the roof. The pitch reduction shortend the span and allowed removal 2 of the purlins and a lot of structural constraints.

[scottishjohn]

4 hours ago, newhome said:

Evac tubes. 

ECO – 58-1800-30 x 2 so 60 tubes

475l TS

Deltasol BS/4 controller. 

 

I can’t go on the roof myself and no one else round here wants to so stalemate. 

 

I only had 2x20 tubes thermomax  --and they boiled 300 litres --so something wrong with your set up if you never got very hot water

[newhome]

1 hour ago, scottishjohn said:

I only had 2x20 tubes thermomax  --and they boiled 300 litres --so something wrong with your set up if you never got very hot water

 

Yes it was likely not set up correctly. From memory the water in the TS used to get to just under 40 degrees. Now it doesn’t get the water hot at all even though the controller says it’s ‘working’, hence thought to be an air lock causing the issue. 

[dnb]

A bit more work using an updated version of my house DHW model:

 

I created 6 versions of the house, with various distributions of PV and ST areas on the roof and ran them in parallel for 500 years using randomly generated hot water usage (according to the model I described in the 1st post) and a model of daily solar availability based on recorded data.

 

Two cases of interest are presented here. The all PV case is for 27 300W PV panels with a nominal efficiency of 18%. The second is 18 PV panels of the same spec, and 9m^2 of evacuated tubes. This should serve to illustrate the case scottishjohn discusses - what do 2 different systems do with the same solar input.

 

I modelled the efficiency of these as 80% when the thermal store (or UVC - it makes no difference to the model) temperature is low, rolling off as an exponential decay as the store temperature increased according to [efficiency = nominal * exp(-tank_temperature/constant)]. I set up the constant to give an efficiency of approximately 25% at 60 deg C tank temperature. I assumed the thermal store had a capacity of 500 litres. Heat losses from the tank are modelled as a basic daily loss, applied as part of the water use calculation.

 

All heating from electricity is assumed to be a very simple immersion heater. I know it can be done more efficiently with ASHP etc, but that adds new variables and more system cost that will be looked at in a later model.

 

Obviously take the results with a pinch of salt until I have properly reviewed the maths. I might have got things wrong.

 

Case 1 required the following yearly energy input costs from the grid (a simple histogram of all 500 years worth of data):

 

Case 2 energy input on the same scale:

 

As can be seen, case 2 requires less energy input, so is in theory is more efficient. But it is only £40/year on average, so when system costs are factored in, case 1 may be the best route. This cost data is available and will be presented later.

 

Now here are some specifics. I have taken one year from the data set and plotted internal data from the model for both cases. It concerns "filling" and "emptying" the tank with energy. The model assumes water is consumed just after 6AM and just after 8PM (It simplifies things nicely) and that once the tank is full the thermal energy can't be used. The PV might be used so I don't count it as waste, but it is not used further because it isn't doing anything for the DHW unless I make the modelled system more complex - and that's tomorrow's job.

 

Case 1

 

Case 2:

 

So on the one-of-one observation it seems the PV system generally wants more energy input throughout the year, but at a low level. Both systems can cope with demand fairly well, so if my model is to be believed, it is a case of finding the cheapest system to install that gives the fewest side effects to the rest of the house rather than DHW running costs.

 

Any thoughts?  Does the model look like it represents reality?

 

One last thought - the wife read the comments about water saving and didn't like the idea of re-education. I pointed her at part G of the building regs. Let's just say it could have gone better.

[Jeremy Harris]

My personal view is that it's best to keep things as simple and reliable as possible, and also to keep the initial capital cost down to as low a level as meets all your requirements.

 

Modelling is great fun (I spent months doing it...), but ultimately there are lots of parameters that are not easy to fully model.  For example, solar thermal hot water can only ever deliver useful energy when the temperature of the collector exceeds the temperature of the hot water storage system.  This means that solar thermal, despite having a high efficiency, can only usefully heat water once the collector has risen to a high enough temperature, and this depends on both the insolation level and the overnight cold soak temperature, so is not easy to model.  What happens in practice is that there is a variable delay in the morning between the sun coming up and the system starting to deliver sensible heat to the hot water system, and that delay is also dependent on the temperature of the hot water storage system at that time (which will vary with pattern of use).

 

By contrast, PV has a lower efficiency, but doesn't care about temperature, if anything it will work more efficiently when the panels are cold first thing in the morning.  PV will always transfer sensible heat to the hot water storage system as long as electricity is being generated and fed to the heating element.  There is no dependence on either the outside temperature or the temperature of the hot water storage system (as long as the storage system thermostat is calling for heat).

 

In terms of cost, then PV is generally a lot cheaper through life than solar thermal, and is maintenance free.  Last time I checked, the installation cost of solar thermal was a lot higher than that for PV, on a Watt-for-Watt comparison basis.  There's also the fact that solar thermal is a one trick pony; all your investment can do is heat hot water.  A PV system can heat hot water, supply power to the house and export power to the grid.