Setting the Scene

When I started designing Tranquility (T) with the intention of making it the lowest energy (financially effective) house I could, I quickly realised that conventional SHW systems didn’t seem to be remotely viable.  Or at least they could not be effective as sold.  Having already concluded solar PV was hopeless this was very disappointing.  However, it was clear that of all the energies we could catch on site – getting the sun to heat our water had logically to have the best potential outcome.  It was time for both detailed analysis and some creative thinking.  The calculations seemed endless, but only a thorough analysis is ever worthwhile.

Starting with the amount of heat energy arriving through each month of the year in the UK (at my latitude) on a 45° plane (that is on a roof pitched at 45°) we have the energy available.  I would like here to publicly thank the MET Office for giving access to substantial amounts of archived data, without which T could not have been as optimally designed as it was.

We will look at some figures later, but it was instantly obvious the amount of hot water that would be provided was not going to make T remotely hot water self sufficient through the summer let alone the winter, if it worked as the sold systems do, but there are two issues:

  1. Is SHW financially viable – that means is it worth investing in OR will we get a return on what it costs to install?
  2. Can it or will it provide most of our hot water?

I then calculated the breakeven price for such a system in order for it to be just economic.  The answer turned out to be £837 assuming it paid for itself in 20 years (the lifetime) and the interest required on the money was 5%.  With systems generally costing umpteen thousands of pounds even with the biggest grants thought of – they cannot be cost effective which answers point 1.

Considering the second point, a raw calculation of the energy available could confuse us because we don’t get exactly the same amount of sunshine each day which means we cannot use the average.  For some days we bake and then for many more we don’t.  And the amount of energy we would need to heat just the water we want fails to account for the energy lost from the storage system, and a multitude of other factors as well.

We will consider some of what was done at Tranquility in the subsequent White Paper (Part 2), but the result is that the cost of supplying T with hot water for the year 2009/10 was just over £7.10 and for 2008/9 – £8.24 which we will ‘prove’ later.  It shows that at least superb hot water performance can be achieved.

But while T is as efficient as any building I know of, there is a maxim “We can (actually) build a zero carbon house – but can the occupiers live in it in a zero carbon way?”  Take a house like T but where all the occupiers take one hour showers or big baths every day; use the kitchen sink with the hot tap left running; leave the windows open through the winter; use the washing machine every day without thought of even the time it is used; set the house temperature at 25° – and you get the picture.  Under these conditions no house can deliver a zero carbon life.  So the people need to work with the building(s) not against them.   Living in T for what is now over 4 years we are unaware of the things we have changed in the way we live, but life is enormously better than it was – not worse.  We live low energy automatically.

System Performance Compared – What do current Installations Achieve?

I believe the information available in T is second to none and you may or may not agree with me but it is what I have to work with and from.  T is a substantial research centre, recording vast amounts of real time data from a big array of sensors – so as you will see it is simple to know precisely how much energy each element of the solar hot water system captures.  This not only allows options to be compared and the optimal option determined, but also means it can be demonstrated or ‘proved’.

By way of demonstration we will look at two perhaps unique charts from the project house:

Sausage Egg & Chips: and no it isn’t a joke

This shows when the frying pan was being heated; when the lid was placed on it; when it was lifted to turn the sausages and put the eggs in, and when it was lifted to serve the cooked food.  But it also shows how long before every trace of cooking smells was gone at each stage.

When trees breathe

This tracks the intensity of CO2 in the outside air across one day.  Fascinatingly it shows when the (mostly) trees start absorbing the carbon in the atmosphere (and therefore emitting oxygen) and when the process reverses.  In everything analysed, the important time is dawn and dusk and not sunrise and sunset. Photosynthesis stopped very close to 8.45pm which is when the CO2 levels started to rise, and the carbon was being absorbed again before 8am.

So the data available is quite comprehensive and accurate.  We are not into guesswork or remote estimations.

We now need to compare two houses:

i.        A conventional one without SHW.

ii.        A house with a normally marketed SHW system.


This White Paper is just that, so not a detailed research report including all the figures and analyses.  It will include sufficient to demonstrate the conclusions.   A full report will be published on this web site which will include all the research and calculations to allow those who take it to scrutinize the work; consider the conclusions and debate them.

The date range covered is from 9th March to 26th July 2009 which covers 20 weeks of what should be almost the most effective period of the year.

In order to discover what proportion of domestic hot water can be produced by a SHW system we need to know the following and only the following:

1. How much hot water does the house use?

2. How much heat is lost from the hot cylinder?

3. How much heat energy does the SHW system collect and when does it collect it?

4. How much fuel does a boiler therefore use to heat the water if there isn’t a SHW system?

5. And how much if there is one?

As T doesn’t have an ‘as sold’ SHW system the data here had to be used to determine how and when a conventional system would work, which isn’t difficult.  A very reputable supplier was contacted and I am very grateful for the information their technical team gave which defined precisely how the systems work.  It did though only confirm my expectations but was useful for that.  T has 4 sq m of flat plates which collect unobstructed sunshine all year from around 10 am.

In order to make a fair comparison the following facts and assumptions were made – but are constant for each option so they are directly comparable:

I.            The amount of hot water being used is taken to be 150 litres/day of hot water at hot tap temperature (delivered from cylinder at 48°).  This is not the same as 150 litres of cylinder water as that might be up to 90° which would be a huge amount of heat and instantly burn.

II.            The cylinder holds 210 litres with the conventional 2 heat exchangers:

  1. The one in the middle is for the boiler and heats just the top 125 litres.
  2. The one close to the bottom is for the solar panel which heats 200 litres.
  3. The bottom 10 litres are heated by downward radiation.

III.            When the water at the top is at or below 48° the boiler will fire and will then heat it through to 60° to reduce the costs and energy losses of boiler cycling.

Notes to explain the work:

  • The ‘hot water cylinder’ in T has 10 sensor pockets enabling detailed temperature data to be recorded.  The sensors in this cylinder record continuously.
  • The sensors are sensitive and measure to 0.1°(C).

How much hot water does the house use?

From the above we are taking hot water use at 150 litres/day which must be heated from the mains at about 8.5° to the delivery temperature of 48°.  The heat energy delivered from the cylinder will therefore be 48.28kw.hrs/week

How much heat is lost from the cylinder?

There is a 140 litre test cylinder installed in T that can be insulated to any specification, and as we see here this enables the precise heat lost from an installed cylinder to be calculated and not estimated.

Plot of cylinder temperatures with an ‘as delivered’ solar cylinder

The insulation was as normal for a solar cylinder – 60mm of Rockwool with a bonded alumised reflective outer layer.  A high specification. 30mm of polystyrene is not better.

The cylinder was heated to 76.0° and monitored continuously for 42 hours.

In this format the cylinder was losing 3.79kw.hrs in 24 hours or 5.51kw.hrs/24 hours for a 210 liter cylinder.

Having examined umpteen charts of cylinder temperatures it is clear just how much heat is lost through the bottom, yet I am not aware of a cylinder that comes with a well insulated base.  High bottom heat losses shouldn’t be surprising as while heat convects upwards, it radiates in every direction equally.

So for the next test the bottom was insulated with about 40mm of fiberglass on average (it has a concave bottom), and two connecting pipes were valved shut to almost stop pipe connection losses. These are two minor improvements yet the Chart below shows the substantial reduction in heat losses.

Plot of cylinder temperatures with improved insulation

Amazing isn’t it?  This is the real difference so we should take manufacturers ‘information’ with caution.  The result is that with this T improved format the Solar Cylinder is losing 2.12kw.hrs for 140 liters or 3.08kw.hrs for 210 liters.  The manufacturers may claim just over 2kw.hrs but this is in perfect laboratory conditions which does not represent your house.  T is extremely good but still these charts apply here.

We are therefore assuming the heat lost from the cylinder is 21.56kw.hrs/week

How much heat energy does the SHW system collect?

The full report shows all the charts but here we will consider just one – though difficult to choose which one!  So we are taking the week from 6th July to 13th July 2009 which should be very hot with lots of hot water.  After all the suppliers tell us their SHW systems can produce “up to 70% of hot water for the whole year so they must be on 100% for the summer.

Plot showing SHW plate collection for week 6th to 13th July 2009

The 3 plots we need to follow are the green, red and blue ones.  The green shows the temperature close to the top of the cylinder; the red is in the middle – exactly at the top of the boiler heat exchanger; and the blue is close to the bottom in the middle of the solar heat exchanger.  Every time the solar plate heats the water the blue plot rises until it reaches the red plot, after which they all rise together.  When the boiler fires the red and blue must rise while leaving the blue behind.  It isn’t possible for the facts to be miss-represented.

[So as not to leave you wondering what the yellow plot is – it is the temperature of the water coming in to C1 to be solar heated].


  • For each day we have the temperature rise at each sensor.  Knowing the volume of water covered by each sensor – we know how many liters x °C or kw.hrs of heat energy were captured.
  • On 12th July a good amount of energy arrived, but there was little to nothing from 6th to the 12th.  Only on two days – the 6th and 10th – was there sufficient to raise the temperature of the bottom 115 liters far enough to have any impact on the top part of the cylinder.  IN JULY.
  • The last time a lot of solar energy arrived before this week was on 2nd July.
  • We can only use these charts to calculate how much solar energy arrived.   Nothing else can be deduced without understanding how the T system works which we will consider later.

The summarized data is as follows, where both the liters x °C and kw.hrs are shown for each week.

Week No Dates Litres x °C KW.hrs
1 9/3 – 15/3 5,895 6.86
2 16/3 – 22/3 17,302 20.14
3 23/3 – 29/3 8,632 10.04
4 30/3 – 5/4 12,825 14.93
5 6/4 – 12/4 10,192 11.86
6 13/4 – 19/4 16,117 18.76
7 20/4 – 26/4 19,252 22.41
8 27/4 – 3/5 17,332 20.18
9 4/5 – 10/5 13,537 15.76
10 11/5 – 17/5 11,700 13.62
11 18/5 – 24/5 18,607 21.66
12 25/5 – 31/5 16,245 18.91
13 1/6 – 7/6 15,060 17.53
14 8/6 – 14/6 15,030 17.50
15 15/6 – 21/6 8,288 9.65
16 22/6 – 28/6 17,137 19.95
17 29/6 – 5/7 13,815 16.08
18 6/7 – 12/7 11,872 13.82
19 13/7 – 19/7 7,792 9.07
20 20/7 – 26/7 13,455 15.66

Note: this data was taken from the charts twice with a few months between them to make certain I did not make a significant mistake.  The two figures differed only by 1.00%.

This gives us an average of 15.71 KW.hrs collected each week through this period which is therefore the amount of heat energy the SHW plates collected

How much fuel does a boiler use to heat the water if there isn’t a SHW system?


The diagram shows the primary heat flows as quantified above:

All figures are in KW.hrs/week

For the period 9/3/09 to 26/7/09

HW = heat delivered from the cylinder

C = Cylinder Heat Losses

B = heat delivered to the cylinder by the boiler

Boiler Energy Burnt = 91.11KW.hrs [see below]

Clearly the boiler must provide the heat delivered plus the heat lost which is the 69.83KW.hrs, but there are other very significant factors to take into account which apply whenever the boiler is used.

When it fires the temperature of the cylinder is already 48°, but the temperature of the boiler itself and the water inside both it and the pipes between the boiler and the cylinder will be in the 18° to 24° range [the boiler is ‘cold’].  The boiler has to fire for some time before it and the water in this circuit is hot enough to start to transfer heat to the water in the cylinder.  The boiler in T is small, but to measure the amount of gas required to achieve the above was easy:

  • The gas meter by the boiler (that only serves the boiler) was read.
  • The live chart showing the water temperatures in the cylinder were put on screen with 10 second data updates.
  • The boiler was fired (done from the computer) and the chart watched.
  • At the moment the temperature sensor showed a rise the boiler was turned off.
  • The boiler gas meter was then read.

In Tranquility it takes 1.28 kw.hrs of gas to be burnt before any heat at all is transferred to the cylinder, so every firing ‘costs’ or ‘wastes’ that 1.28 kw.hrs.

Of course there are pipe losses between the boiler and cylinder which would make it worse, but it is assumed here that the boiler is operating at an efficiency of 85% and that all that heat is transferred to the cylinder!  Tranquility does not have, nor would I fit, a condensing boiler but more of that in another document.

The amount of heat energy required to arrive at the cylinder is therefore the 69.83 kw.hrs/week but:

  1. The 69.83kw.hrs becomes 82.15 kw.hrs of gas after allowing for the boiler inefficiency.
  2. Examining the temperatures and the hot water flows, the boiler would be required to fire essentially every day and once per day. Each week therefore 7 times the 1.28kw.hrs must be added.
  3. 3. The energy burnt in the boiler is therefore 91.11 kw.hrs.

How much fuel does a boiler use to heat the water if there is a SHW system?

The diagram shows the primary heat flows now including the Solar:

All figures are in KW.hrs/week

For the period 9/3/09 to 26/7/09

HW = heat delivered from the cylinder

C = Cylinder Heat Losses

B = heat delivered to the cylinder by the boiler

Boiler Energy Burnt = 67.51KW.hrs [see below]

The amount of heat energy the boiler must now provide at the cylinder is: the heat delivered plus the heat lost minus the heat gained from the solar panels which is now 54.12kw.hrs but:

  1. Taking the same boiler inefficiency this becomes 63.67kw.hrs.
  2. During the 20 weeks or 140 days there was no effective solar energy gained on 42 days, and on a further 18 days insufficient for the water in the top of the cylinder to be heated at all.  So the boiler would still have to fire on 60 of those days and if you work through the water flows it would be many more.  But taking the 60 days the total gas energy burnt by the boiler will be 67.51 kw.hrs.

Analysis and conclusions:

I often say “Wanting something to work doesn’t make it do so”.

The amount of gas energy burnt by the boiler is therefore reduced from 91.11 to 67.51 kw.hrs by having the SHW system through this summer period.  So it reduced the energy bills for the householder by just 26%.  You might take the view this is pitiful as I do, and you might be forgiven for feeling those who sell these systems don’t just oversell them but clearly live on a different planet.  If it wasn’t so serious I might suggest they would work well on Venus but not here.

Owners of such systems have no way of knowing what they generate.  They have no data like Tranquility to demonstrate and prove what is collected.  They just trust they are doing well and having spent a fortune on them are most unwilling to accept they don’t really work.

If we take the analysis above, the SHW system would reduce the gas used by 27.44 kw.hrs/week for the 20 weeks which is less than a £20 saving.  Double it for the rest of the summer and the figure is way less than £40.  Now given most of these cost around £5,000 installed, are we happy with a £40 return for that investment?

It is not therefore surprising that some have pursued the suppliers through the courts – and won.

It is absolutely impossible no matter what the weather for any of these systems to provide the ‘Up to a 70% saving’ that is so often claimed.  Government has a big responsibility here as I receive many such marketing approaches through the post which start ‘Government Approved’.  If I was government I would run a mile from such claims.

Mike Hillard.

A further White Paper will be published which explains some of the reasons why the Tranquility system actually does work and works well, and it will also include advice on how to very cost effectively better insulate any cylinder you have.