Wanting to reduce gas usage - New Heating Design Circuit - will it work?

70sbudgie

Solarchaser said:

With that in mind, I'd say that if you spec a custom build with two coils, you negate the need for two plate heat exchangers and two pumps, which I'd imagine will end up around cost neutral (additional cost of custom tank) when you add it all together, but will be a simpler system.

If you then have the cold and hot inlets/outlets plumbed to your system boiler, you have your heating circuit without a 3rd coil.

On the flow detector, I haven't got mine to work.
It's essentially a paddle in the water flow, but the truth is that most of the time running a tap doesn't actually use very much water,  most tap hoses have an inner diameter of no more than 8mm, so trying to detect that flow in a 22mm pipe didn't work, and in a 15mm pipe didn't really work either. If I could convince you to remove the loop, then there is no need for the sensor either. But obviously it's your system and you need to decide how important instant hot water is.

Is it still possible to buy "off the shelf" dual core cylinders? What would be the benefit of a custom build?

I am very interested why @Solarchaser seems to prefer coil heat exchangers rather than plate heat exchangers? Can you explain a bit more? 

I am also trying to get my head around how a dual core cylinder would change the diagram above - would it be one core for DHW (essentially an open circuit) and the other for the central heating (a closed circuit)? The heated water from the boiler on a closed circuit using the hot and cold feed ports on the tank. That would mean you'd have approx 500l circulating through your boiler? 

Freepost said:

2) Why a 500l tank, and why horizontal - My current gas consumption in the depths of winter (January) is around 70kW/day. With 2 6kW immersion heaters running for a 4hr period I can get near enough 50kW into the tank and the gas boiler will have to make up the short fall until I install batteries. Further to this and to give some idea about my arithmetic I used a formula of Volume of tank x 4 x temperature rise / 3412 = Power (kW) required from https://www.boilerguide.co.uk/articles/immersion-heaters where the tank is 500l and my available energy in a four hour period from the immersion heater is 48kw then a little bit of algebra delivers approx 80°C. .... The tank will be mounted vertically.

To me, this looks like an 80°C temperature rise? This seems a lot and might be problematic in the summer, when the cold water feed might be 15°C, or are you just calculating for the max and expect to modulate?

There also seems something not quite right about your units. You mention 48kW in a 4 hour period as the energy available. I think you mean 48kWhrs? If you were to switch to the 3kW + 6kW immersion heater as @Solarchaser suggested to utilise solar, then when doing so, you would be heating at a different rate. You may find that even in winter, a couple kWhrs from solar will be enough to get you closer to the 80° temp rise. I don't imagine there will be too many days when the 36kWhrs from Go wouldn't be enough. And its not like you can't do a short peak rate or gas boost. And don't forget that the temp rise is on the cold feed water temperature, not the ambient air temperature - there is less range on the former.

My gut feel is that your calculations have oversized your required capacity, but I don't know if that is because I am struggling to follow all the detail?

michaels

Ectophile said:

Your heat exchanger for the DHW is going to have to be really good if you want anything other than vaguely warm water coming out of the taps.  Unless your system is really well insulated, and you're going to circulate it round and round until it eventually reaches an acceptable temperature.

I had a set up like this in my old house and there was no problem with the plate heat exchanger heating the DHW supply at mains pressure and volumes, probably the best hot water supply I have had anywhere.   May have had a themomixer after the heat exchanger on the hot water supply to blend in cold to get the right temp, didn't have a return loop but it was only a 3 bed terrace so the runs to taps were not excessive.

Due to limited loft space we had a custom horizontal tank with various devices in place to avoid mixing to maintain a 'hot top' for the how water supply  Think the tank was direct to the gas boiler with a separate heating loop also direct to the tank with two draw offs (top and bottom) and a thermomixer to give control over the temp of the water going to the rads.  Did have an immersion but it was for backup not TOU.

So definitely it will all work, just need to be clever with the mixing.  Not sure if you need the second heat exchanger or can just use the tank flow for the rads.

Solarchaser

70sbudgie said:

Solarchaser said:

With that in mind, I'd say that if you spec a custom build with two coils, you negate the need for two plate heat exchangers and two pumps, which I'd imagine will end up around cost neutral (additional cost of custom tank) when you add it all together, but will be a simpler system.

If you then have the cold and hot inlets/outlets plumbed to your system boiler, you have your heating circuit without a 3rd coil.

On the flow detector, I haven't got mine to work.
It's essentially a paddle in the water flow, but the truth is that most of the time running a tap doesn't actually use very much water,  most tap hoses have an inner diameter of no more than 8mm, so trying to detect that flow in a 22mm pipe didn't work, and in a 15mm pipe didn't really work either. If I could convince you to remove the loop, then there is no need for the sensor either. But obviously it's your system and you need to decide how important instant hot water is.

Is it still possible to buy "off the shelf" dual core cylinders? What would be the benefit of a custom build?

I am very interested why @Solarchaser seems to prefer coil heat exchangers rather than plate heat exchangers? Can you explain a bit more? 

I am also trying to get my head around how a dual core cylinder would change the diagram above - would it be one core for DHW (essentially an open circuit) and the other for the central heating (a closed circuit)? The heated water from the boiler on a closed circuit using the hot and cold feed ports on the tank. That would mean you'd have approx 500l circulating through your boiler? 

Freepost said:

2) Why a 500l tank, and why horizontal - My current gas consumption in the depths of winter (January) is around 70kW/day. With 2 6kW immersion heaters running for a 4hr period I can get near enough 50kW into the tank and the gas boiler will have to make up the short fall until I install batteries. Further to this and to give some idea about my arithmetic I used a formula of Volume of tank x 4 x temperature rise / 3412 = Power (kW) required from https://www.boilerguide.co.uk/articles/immersion-heaters where the tank is 500l and my available energy in a four hour period from the immersion heater is 48kw then a little bit of algebra delivers approx 80°C. .... The tank will be mounted vertically.

To me, this looks like an 80°C temperature rise? This seems a lot and might be problematic in the summer, when the cold water feed might be 15°C, or are you just calculating for the max and expect to modulate?

There also seems something not quite right about your units. You mention 48kW in a 4 hour period as the energy available. I think you mean 48kWhrs? If you were to switch to the 3kW + 6kW immersion heater as @Solarchaser suggested to utilise solar, then when doing so, you would be heating at a different rate. You may find that even in winter, a couple kWhrs from solar will be enough to get you closer to the 80° temp rise. I don't imagine there will be too many days when the 36kWhrs from Go wouldn't be enough. And its not like you can't do a short peak rate or gas boost. And don't forget that the temp rise is on the cold feed water temperature, not the ambient air temperature - there is less range on the former.

My gut feel is that your calculations have oversized your required capacity, but I don't know if that is because I am struggling to follow all the detail?

It's definitely still possible to buy off the shelf dual coil cylinders, but its what they will be used for.
Generally coils in tanks are to provide heat to the tank from a back boiler, stove, heat pump etc, and so their position in the tank will be fairly low down, the cooler end of the tank.

Because freepost is essentially using the cylinder in reverse (as am i) the goal would be to have the coils further up to extract as much heat as possible. 

Why coils instead of flat plate, simply, simplicity. 
If you have flat plates you have 4 connections to a water body outside of the cylinder, but more importantly you need a pump to pump the hot fluid from the cylinder to the flat plate, with a coil it simply flows inside either by mains water pressure, or by the central heating pump.
Less pumps, less bits to break down.
Less connections, less bits to leak.

Your thoughts are correct, 1 coil open circuit to provide heat to the water turning cold to hot potable water, the other coil would be closed circuit for space heating.
Full disclosure this is how my cylinders are working, which is probably why I favour it.
Feed and supply would be back to boiler making a 500l boiler circuit. (Though I'd definitely have full bore valves on the in and out to be able to isolate the circuits from each other.)

Good spot on the heat rates, I never even noticed that.
General rule of thumb is 1 kw for 1 hour in 1000l is 1C rise.
So 1kw in 500L is 2C rise.

So really if we assume a 40C normal lowest temperature for the tank, as under that won't provide hot water or much space heating, and say a 85C max, then the delta is 45c, so half of that for the 500L tank gives 22.5Kwh to heat the tank each winters day, over 4 hours its just over 5kw/hour, so really 2x3kw elements will be about ideal.
Truth be told I've done these calculations before, as I have a 500L tank.

For clarity I have 2x 3kw and a 3 hour go faster period, so I'll get 18kwh or 36C max..... actually not really as I have another 300l tank with another 2x3kw elements so I'll get more like 36kwh.

Now in truth the 1kw /1hr /1000l =1C isn't really true, and so the 22.5kwh isn't really true. 
The truth is more like this

So 26kw/4 hours is 6.5kw, so the 2x3kw is slightly under, but its pretty darn close.

Edit. Missed the question about coils vs plates

michaels

Pretty sure we had single plate for hot water and no coil, just direction connections to separate boiler and heating circuits so perhaps 3 pumps.  Both boiler and hot water circuits used blender valves to try and send water to rads at desired temp even when tank was hotter and make return to boiler temp so as to maximise condensing.

zeupater

Freepost said:

Hello,

I'm looking at installing a new central heating system with UHF on the ground floor and normal rads on the 1st Floor.

[...]

... I'm considering changing/modifying this design to use cheap rate electricity from Octopus Go/Agile to heat up the cylinder during the night and/or solar PV in the summer months with the System Boiler giving back up in perhaps the depths of winter. Octopus Go/Agile provides very cheap rate electricity for a 4 to 6 hour window in the night and, although not now, has in the past paid the consumer to take the energy during that period. I'm sure when we get through the present difficulties those days of cheap rate will return.

[...]

Operation;

The water in the Telford Tempest 500l Cylinder is heated by 2 x 3kw Immersion Heaters using Octopus Go/Octopus Agile cheap rate tariff, Energy from the Solar PV System and the Ideal System Boiler ...

[...]

Is this an idea worth pursuing and will it work?

Your thoughts would be most welcome;

F.

Hi

A very important consideration would be your expectations for the future of 'cheap rate' overnight electricity being available at a level of differential which supports the logic you've used.

Overnight electricity tariffs such as those offered by Octopus aren't actually a new concept but date back around half-a-century to the likes of E7 encouraging the use of electric storage heaters in order to to 'mop up' surplus overnight generating capacity, ie to close the supply/demand gap, which due to the popularity of storage heating (bulky/inflexibility) never made significant progress thus leaving considerable overnight generation capacity  .... however, move on to current date we're starting to see the possibility of that gap actually narrowing and potentially closing altogether.

So what's likely fill the gap? ... well, overnight EV charging, domestic battery overnight charging for use the following day, the installation of heat-pumps to replace GCH having longer run times would constitute some major usage profile shifting and when the gap is filled, what's the future for the carrot approach (cheap overnight electricity) and will it logically & simply be superseded by the big stick approach of heavily loading tariffs at peak usage times ...

Looking at your proposal it seems that you'll be investing in a short term solution ... so your 'bet' is effectively one of how short-a-term an investment in a low efficiency electric heating solution will be before your energy costs deem investment in an alternative replacement heat source ... so effectively it's a case of where on the S curve we currently are on domestic batteries, electric vehicles & heat pumps and therefore how long it will be before overnight and standard daytime tariffs equalise, as they are almost certainly to do.

Just food for thought

HTH - Z        

michaels

zeupater said:

Freepost said:

Hello,

I'm looking at installing a new central heating system with UHF on the ground floor and normal rads on the 1st Floor.

[...]

... I'm considering changing/modifying this design to use cheap rate electricity from Octopus Go/Agile to heat up the cylinder during the night and/or solar PV in the summer months with the System Boiler giving back up in perhaps the depths of winter. Octopus Go/Agile provides very cheap rate electricity for a 4 to 6 hour window in the night and, although not now, has in the past paid the consumer to take the energy during that period. I'm sure when we get through the present difficulties those days of cheap rate will return.

[...]

Operation;

The water in the Telford Tempest 500l Cylinder is heated by 2 x 3kw Immersion Heaters using Octopus Go/Octopus Agile cheap rate tariff, Energy from the Solar PV System and the Ideal System Boiler ...

[...]

Is this an idea worth pursuing and will it work?

Your thoughts would be most welcome;

F.

Hi

A very important consideration would be your expectations for the future of 'cheap rate' overnight electricity being available at a level of differential which supports the logic you've used.

Overnight electricity tariffs such as those offered by Octopus aren't actually a new concept but date back around half-a-century to the likes of E7 encouraging the use of electric storage heaters in order to to 'mop up' surplus overnight generating capacity, ie to close the supply/demand gap, which due to the popularity of storage heating (bulky/inflexibility) never made significant progress thus leaving considerable overnight generation capacity  .... however, move on to current date we're starting to see the possibility of that gap actually narrowing and potentially closing altogether.

So what's likely fill the gap? ... well, overnight EV charging, domestic battery overnight charging for use the following day, the installation of heat-pumps to replace GCH having longer run times would constitute some major usage profile shifting and when the gap is filled, what's the future for the carrot approach (cheap overnight electricity) and will it logically & simply be superseded by the big stick approach of heavily loading tariffs at peak usage times ...

Looking at your proposal it seems that you'll be investing in a short term solution ... so your 'bet' is effectively one of how short-a-term an investment in a low efficiency electric heating solution will be before your energy costs deem investment in an alternative replacement heat source ... so effectively it's a case of where on the S curve we currently are on domestic batteries, electric vehicles & heat pumps and therefore how long it will be before overnight and standard daytime tariffs equalise, as they are almost certainly to do.

Just food for thought

HTH - Z        

Very good point and part of the reason I decided to get a second Nissan Leaf to do V2H rather than a battery system - if prices go back to the old normal I can simply sell the car on and recoup the capital.  I am slso thinking of adding a second hot water cylinder as it seems like at current prices I could produce enough hot water during the 4.5p window to meet our entire daily usage at 4.5p per kwh compared to gas that was gong to cost 14p increasing to possibly as much as 20p which should pay for the cylinder in 2 years - but the price freeze has already changed the maths and no doubt it may change again...

70sbudgie

zeupater said:

Freepost said:

Hello,

I'm looking at installing a new central heating system with UHF on the ground floor and normal rads on the 1st Floor.

[...]

... I'm considering changing/modifying this design to use cheap rate electricity from Octopus Go/Agile to heat up the cylinder during the night and/or solar PV in the summer months with the System Boiler giving back up in perhaps the depths of winter. Octopus Go/Agile provides very cheap rate electricity for a 4 to 6 hour window in the night and, although not now, has in the past paid the consumer to take the energy during that period. I'm sure when we get through the present difficulties those days of cheap rate will return.

[...]

Operation;

The water in the Telford Tempest 500l Cylinder is heated by 2 x 3kw Immersion Heaters using Octopus Go/Octopus Agile cheap rate tariff, Energy from the Solar PV System and the Ideal System Boiler ...

[...]

Is this an idea worth pursuing and will it work?

Your thoughts would be most welcome;

F.

Hi

A very important consideration would be your expectations for the future of 'cheap rate' overnight electricity being available at a level of differential which supports the logic you've used.

Overnight electricity tariffs such as those offered by Octopus aren't actually a new concept but date back around half-a-century to the likes of E7 encouraging the use of electric storage heaters in order to to 'mop up' surplus overnight generating capacity, ie to close the supply/demand gap, which due to the popularity of storage heating (bulky/inflexibility) never made significant progress thus leaving considerable overnight generation capacity  .... however, move on to current date we're starting to see the possibility of that gap actually narrowing and potentially closing altogether.

So what's likely fill the gap? ... well, overnight EV charging, domestic battery overnight charging for use the following day, the installation of heat-pumps to replace GCH having longer run times would constitute some major usage profile shifting and when the gap is filled, what's the future for the carrot approach (cheap overnight electricity) and will it logically & simply be superseded by the big stick approach of heavily loading tariffs at peak usage times ...

Looking at your proposal it seems that you'll be investing in a short term solution ... so your 'bet' is effectively one of how short-a-term an investment in a low efficiency electric heating solution will be before your energy costs deem investment in an alternative replacement heat source ... so effectively it's a case of where on the S curve we currently are on domestic batteries, electric vehicles & heat pumps and therefore how long it will be before overnight and standard daytime tariffs equalise, as they are almost certainly to do.

Just food for thought

HTH - Z        

What sort of timescales are you thinking of? I received my E7 October rates from Octopus at the weekend and was interested to see the day/night differential actually increase. I'm not convinced that the investment in the UK electricity network over the next 10 years is going to be suffient to remove the day / night differential. And if if it does, I think it will be replaced by a mix of different time of use tariffs. So having a heating system that can be flexible about charging / discharging will be beneficial. 

Perhaps it might be worth considering how to future proof the system - could an A2WHP replace the boiler?

70sbudgie

Solarchaser said:

70sbudgie said:

Solarchaser said:

With that in mind, I'd say that if you spec a custom build with two coils, you negate the need for two plate heat exchangers and two pumps, which I'd imagine will end up around cost neutral (additional cost of custom tank) when you add it all together, but will be a simpler system.

If you then have the cold and hot inlets/outlets plumbed to your system boiler, you have your heating circuit without a 3rd coil.

On the flow detector, I haven't got mine to work.
It's essentially a paddle in the water flow, but the truth is that most of the time running a tap doesn't actually use very much water,  most tap hoses have an inner diameter of no more than 8mm, so trying to detect that flow in a 22mm pipe didn't work, and in a 15mm pipe didn't really work either. If I could convince you to remove the loop, then there is no need for the sensor either. But obviously it's your system and you need to decide how important instant hot water is.

Is it still possible to buy "off the shelf" dual core cylinders? What would be the benefit of a custom build?

I am very interested why @Solarchaser seems to prefer coil heat exchangers rather than plate heat exchangers? Can you explain a bit more? 

I am also trying to get my head around how a dual core cylinder would change the diagram above - would it be one core for DHW (essentially an open circuit) and the other for the central heating (a closed circuit)? The heated water from the boiler on a closed circuit using the hot and cold feed ports on the tank. That would mean you'd have approx 500l circulating through your boiler? 

Freepost said:

2) Why a 500l tank, and why horizontal - My current gas consumption in the depths of winter (January) is around 70kW/day. With 2 6kW immersion heaters running for a 4hr period I can get near enough 50kW into the tank and the gas boiler will have to make up the short fall until I install batteries. Further to this and to give some idea about my arithmetic I used a formula of Volume of tank x 4 x temperature rise / 3412 = Power (kW) required from https://www.boilerguide.co.uk/articles/immersion-heaters where the tank is 500l and my available energy in a four hour period from the immersion heater is 48kw then a little bit of algebra delivers approx 80°C. .... The tank will be mounted vertically.

To me, this looks like an 80°C temperature rise? This seems a lot and might be problematic in the summer, when the cold water feed might be 15°C, or are you just calculating for the max and expect to modulate?

There also seems something not quite right about your units. You mention 48kW in a 4 hour period as the energy available. I think you mean 48kWhrs? If you were to switch to the 3kW + 6kW immersion heater as @Solarchaser suggested to utilise solar, then when doing so, you would be heating at a different rate. You may find that even in winter, a couple kWhrs from solar will be enough to get you closer to the 80° temp rise. I don't imagine there will be too many days when the 36kWhrs from Go wouldn't be enough. And its not like you can't do a short peak rate or gas boost. And don't forget that the temp rise is on the cold feed water temperature, not the ambient air temperature - there is less range on the former.

My gut feel is that your calculations have oversized your required capacity, but I don't know if that is because I am struggling to follow all the detail?

It's definitely still possible to buy off the shelf dual coil cylinders, but its what they will be used for.
Generally coils in tanks are to provide heat to the tank from a back boiler, stove, heat pump etc, and so their position in the tank will be fairly low down, the cooler end of the tank.

Because freepost is essentially using the cylinder in reverse (as am i) the goal would be to have the coils further up to extract as much heat as possible. 

Why coils instead of flat plate, simply, simplicity. 
If you have flat plates you have 4 connections to a water body outside of the cylinder, but more importantly you need a pump to pump the hot fluid from the cylinder to the flat plate, with a coil it simply flows inside either by mains water pressure, or by the central heating pump.
Less pumps, less bits to break down.
Less connections, less bits to leak.

Your thoughts are correct, 1 coil open circuit to provide heat to the water turning cold to hot potable water, the other coil would be closed circuit for space heating.
Full disclosure this is how my cylinders are working, which is probably why I favour it.
Feed and supply would be back to boiler making a 500l boiler circuit. (Though I'd definitely have full bore valves on the in and out to be able to isolate the circuits from each other.)

Good spot on the heat rates, I never even noticed that.
General rule of thumb is 1 kw for 1 hour in 1000l is 1C rise.
So 1kw in 500L is 2C rise.

So really if we assume a 40C normal lowest temperature for the tank, as under that won't provide hot water or much space heating, and say a 85C max, then the delta is 45c, so half of that for the 500L tank gives 22.5Kwh to heat the tank each winters day, over 4 hours its just over 5kw/hour, so really 2x3kw elements will be about ideal.
Truth be told I've done these calculations before, as I have a 500L tank.

For clarity I have 2x 3kw and a 3 hour go faster period, so I'll get 18kwh or 36C max..... actually not really as I have another 300l tank with another 2x3kw elements so I'll get more like 36kwh.

Now in truth the 1kw /1hr /1000l =1C isn't really true, and so the 22.5kwh isn't really true. 
The truth is more like this

So 26kw/4 hours is 6.5kw, so the 2x3kw is slightly under, but its pretty darn close.

Edit. Missed the question about coils vs plates

Why 40°C starting temp? That won't be the temperature of the cold water feed.

QrizB

70sbudgie said:

Why 40°C starting temp? That won't be the temperature of the cold water feed.

Because for a thermal store you're not starting with out-of-the-ground water, you're re-heating the water that you heated yesterday. And (unless you're feeding it to a heat pump) you won't want the thermal store to get any colder than the coldest acceptable DHW temperature.

Solarchaser

70sbudgie said:

Solarchaser said:

70sbudgie said:

Solarchaser said:

With that in mind, I'd say that if you spec a custom build with two coils, you negate the need for two plate heat exchangers and two pumps, which I'd imagine will end up around cost neutral (additional cost of custom tank) when you add it all together, but will be a simpler system.

If you then have the cold and hot inlets/outlets plumbed to your system boiler, you have your heating circuit without a 3rd coil.

On the flow detector, I haven't got mine to work.
It's essentially a paddle in the water flow, but the truth is that most of the time running a tap doesn't actually use very much water,  most tap hoses have an inner diameter of no more than 8mm, so trying to detect that flow in a 22mm pipe didn't work, and in a 15mm pipe didn't really work either. If I could convince you to remove the loop, then there is no need for the sensor either. But obviously it's your system and you need to decide how important instant hot water is.

Is it still possible to buy "off the shelf" dual core cylinders? What would be the benefit of a custom build?

I am very interested why @Solarchaser seems to prefer coil heat exchangers rather than plate heat exchangers? Can you explain a bit more? 

I am also trying to get my head around how a dual core cylinder would change the diagram above - would it be one core for DHW (essentially an open circuit) and the other for the central heating (a closed circuit)? The heated water from the boiler on a closed circuit using the hot and cold feed ports on the tank. That would mean you'd have approx 500l circulating through your boiler? 

Freepost said:

2) Why a 500l tank, and why horizontal - My current gas consumption in the depths of winter (January) is around 70kW/day. With 2 6kW immersion heaters running for a 4hr period I can get near enough 50kW into the tank and the gas boiler will have to make up the short fall until I install batteries. Further to this and to give some idea about my arithmetic I used a formula of Volume of tank x 4 x temperature rise / 3412 = Power (kW) required from https://www.boilerguide.co.uk/articles/immersion-heaters where the tank is 500l and my available energy in a four hour period from the immersion heater is 48kw then a little bit of algebra delivers approx 80°C. .... The tank will be mounted vertically.

To me, this looks like an 80°C temperature rise? This seems a lot and might be problematic in the summer, when the cold water feed might be 15°C, or are you just calculating for the max and expect to modulate?

There also seems something not quite right about your units. You mention 48kW in a 4 hour period as the energy available. I think you mean 48kWhrs? If you were to switch to the 3kW + 6kW immersion heater as @Solarchaser suggested to utilise solar, then when doing so, you would be heating at a different rate. You may find that even in winter, a couple kWhrs from solar will be enough to get you closer to the 80° temp rise. I don't imagine there will be too many days when the 36kWhrs from Go wouldn't be enough. And its not like you can't do a short peak rate or gas boost. And don't forget that the temp rise is on the cold feed water temperature, not the ambient air temperature - there is less range on the former.

My gut feel is that your calculations have oversized your required capacity, but I don't know if that is because I am struggling to follow all the detail?

It's definitely still possible to buy off the shelf dual coil cylinders, but its what they will be used for.
Generally coils in tanks are to provide heat to the tank from a back boiler, stove, heat pump etc, and so their position in the tank will be fairly low down, the cooler end of the tank.

Because freepost is essentially using the cylinder in reverse (as am i) the goal would be to have the coils further up to extract as much heat as possible. 

Why coils instead of flat plate, simply, simplicity. 
If you have flat plates you have 4 connections to a water body outside of the cylinder, but more importantly you need a pump to pump the hot fluid from the cylinder to the flat plate, with a coil it simply flows inside either by mains water pressure, or by the central heating pump.
Less pumps, less bits to break down.
Less connections, less bits to leak.

Your thoughts are correct, 1 coil open circuit to provide heat to the water turning cold to hot potable water, the other coil would be closed circuit for space heating.
Full disclosure this is how my cylinders are working, which is probably why I favour it.
Feed and supply would be back to boiler making a 500l boiler circuit. (Though I'd definitely have full bore valves on the in and out to be able to isolate the circuits from each other.)

Good spot on the heat rates, I never even noticed that.
General rule of thumb is 1 kw for 1 hour in 1000l is 1C rise.
So 1kw in 500L is 2C rise.

So really if we assume a 40C normal lowest temperature for the tank, as under that won't provide hot water or much space heating, and say a 85C max, then the delta is 45c, so half of that for the 500L tank gives 22.5Kwh to heat the tank each winters day, over 4 hours its just over 5kw/hour, so really 2x3kw elements will be about ideal.
Truth be told I've done these calculations before, as I have a 500L tank.

For clarity I have 2x 3kw and a 3 hour go faster period, so I'll get 18kwh or 36C max..... actually not really as I have another 300l tank with another 2x3kw elements so I'll get more like 36kwh.

Now in truth the 1kw /1hr /1000l =1C isn't really true, and so the 22.5kwh isn't really true. 
The truth is more like this

So 26kw/4 hours is 6.5kw, so the 2x3kw is slightly under, but its pretty darn close.

Edit. Missed the question about coils vs plates

Why 40°C starting temp? That won't be the temperature of the cold water feed.

Because there will be no benefit of having it lower either for hot water or heating,  so it wouldn't continue to be used past that, and as QrizB says, this is the point you would start with the water from the day before, rather than new hot water, it's reusing same water and reheating same water.