Well I have a new HW tank to fit soon but I've been trying to optimise the HW a little more - Average consumption is 4kWh per day to do the water heating for two people occupancy.

 

Flow temp is typically 66 Deg C return temp peaks at 61 Deg C, 115 Litre tank and 30 Min Cycle with the tank ending up at 52 Deg C (I've tried the lower and slower approach and to be honest it actually uses more gas but I think that's because of overshooting the target temp when I did it)

 

However because of the X plan set up it does mean even in the summer I put a slug of 68 deg water into the rads straight after the HW cycle is finished. In winter no issues - it helps make up for the HW cycle time out of CH - summer it contribute to heat gain in the house......

 

I decided to do an experiment..........

 

If I run the HW for 30 mins and ten kill the boiler and continue to circulate the HW system with no boiler activity I can get an increase in the tank temp of 2 deg C and the water hitting the rads is down to 50 deg C

 

I'm now thinking about putting a delay timer on the HW valve closing 10 mins after HW cycle is completed to extract the last bit of energy out of boiler and circuit

 

Mad idea?

New Tank Fitted over the weekend

Grant Tank - 180L 2.7m2 coil surface area 

 

 

couple of errors in pipework configuration that are resulting in increased heat loss but I'll fix them over the next couple of weeks.

 

meantime a quick experiment on boiler efficiency (compared to old tank data) revealed

 

Boiler efficiency

 

Old tank when heating HW - efficiency was 88.97% (condensate volume 157 grams - condensate stopped after 10 mins of a 30 min cycle due to return temp being above 54 deg)

 

New tank when heating HW - efficiency is 93.29% (Condensate volume 332 grams was generating condensate all the way thro the 45 min cycle)

 

Can show workings if required....... 

 

Heat Loss

 

Old Hot water tank fully lagged pipework - london loops everywhere (U loops on any hot outlets from tank) best I could get in term of temp decay in the tank was 0.4 deg per hour (most of the time it was 0.5 to 0.6 deg per hour) so heat a tank to 50 deg and 20 hours later it would be down to 40 deg (thats a just bearable shower according to SWMBO)

 

New Tank - no lagging on pipework yet and minimal london loops and the tank losses are down to 0.19 deg per hour - so heat a tank to 50 deg and 24 hours later it's still over 45 deg - I think I need a better hobby!!!!

 

PS Yes I need to fit a tank stat

 

PPS The tank is set up as a vented tank (so no G3 certification is required) 

Was having a conversation with a mate on whatsapp and he decided to put part of the comments above into AI

 

Yes — those numbers suggest the new cylinder is substantially better, not just marginally better. The nice thing is that his temperature-decay measurements let us put sensible numbers on it.
For a hot-water cylinder, a first approximation is
\[
Q = m c_p \Delta T
\]with water \(c_p \approx 4.186\text{ kJ/kg·K}\). So, for every 100 litres of water, a cooling rate of 1 °C/hour corresponds to about:
\[
\frac{100\times4.186}{3.6}=116.3\text{ W}
\]Therefore, at his measured rates:
Cooling rate    Heat loss per 100 L    200 L cylinder equivalent
0.60 °C/h    69.8 W    140 W
0.50 °C/h    58.1 W    116 W
0.40 °C/h    46.5 W    93 W
0.19 °C/h    22.1 W    44 W

So if, for example, these were roughly 200 L cylinders, he's effectively gone from somewhere around 90–140 W of standing heat loss to around 44 W at that part of the temperature range.
That's roughly a 60–65% reduction in heat loss, based simply on the measured cooling rate:
\[
1-\frac{0.19}{0.50}\approx62\%
\]And it may be nearer 68% if the old cylinder was normally doing 0.6 °C/h.
There is another interesting way of looking at his actual example. If the old tank went from 50 → 40 °C in 20 h, a 200 L tank lost approximately:
\[
200\times4.186\times10 = 8372\text{ kJ}
\]or:
\[
\approx2.33\text{ kWh}
\]So that's about 2.3 kWh of stored hot-water energy disappearing every 20 hours.
With the new tank at 0.19 °C/h, after 24 h:
\[
50-(24\times0.19)=45.44^\circ C
\]A 200 L tank has then lost:
\[
200\times4.186\times4.56
\]≈ 1.06 kWh.
So you're comparing roughly:
old: ~2.3 kWh lost in 20 h
versus
new: ~1.1 kWh lost in 24 h
That's a pretty striking difference.
There's one subtlety, though. Heat loss isn't actually constant in °C/hour. It's approximately proportional to the temperature difference between the cylinder and the room:
\[
\dot Q = UA(T_{tank}-T_{room})
\]so the temperature ought to decay exponentially rather than in a perfectly straight line. At 50 °C in a 20 °C room, there's a 30 °C temperature difference; at 40 °C there's only 20 °C. Consequently the old cylinder losing 0.5–0.6 °C/h near 50 °C is actually worse thermally than that simple 20-hour average makes it look.
The London loops are relevant too. They're intended to reduce thermosyphon circulation along hot-water pipes — hot water rising out of the cylinder pipework, cooling, and creating a slow natural circulation loop. But if the new cylinder manages 0.19 °C/h even before the associated pipework has been fully insulated, that's particularly impressive. It suggests most of the improvement is inherent to the cylinder itself: better factory insulation, fewer thermal bridges, improved penetrations/fittings, etc., rather than just clever pipework.
I'd be quite interested to see what happens after he lags all the new pipework. If he got down to, say, 0.15 °C/h, a 200 L cylinder would be losing only about 35 W at that operating point.
So I'd reply to him something along the lines of: your new hobby may be slightly obsessive, but unfortunately your measurements say it's working — you've apparently cut the standing heat loss by something like two-thirds. 😄
If you know the actual cylinder volume, I can calculate the losses in watts, kWh/day and £/year quite accurately from his figures.

 

the bit in bold made us both laugh

Grant documentation says 1.61 kWh over 24hrs and energy lost at a rate of 67 W

Next job is to understand how to wire this into the system

 

 

As a temporary upper limit safety measure I'm using the original tank stat as a pipestat on the coil return

 

 

it's not resting there it's got a spring wire underneath to hold it on 🙂

7 hours ago, marshian said:

Next job is to understand how to wire this into the system

The right is the thermostat and left is overheat protection required for an unvented cylinder - so you only need bother with the right hand side. 

3 hours ago, JohnMo said:

The right is the thermostat and left is overheat protection required for an unvented cylinder - so you only need bother with the right hand side. 

 

Thank You - makes life a little simpler