UFH, Heat pumps and buffer tanks

[ProDave]

So I'm starting to plan installing (at least part of) my heating system.

 

I have an air source heat pump. I am having wet under floor heating upstairs in just the bathrooms, and downstairs in most rooms. Upstairs UFH is installed, downstairs will come later.

 

I am planning a buffer tank of some sort for the heating, and this is where I need help and advice,

 

My understanding is the heat pump will heat the buffer tank, and the UFH will take heat from the buffer tank. That seems simple. But it raises some questions:

 

Firstly just a dumb tank or a tank with heat exchange coils? 

 

On the face of it, a dumb tank would do, but then the entire system, including the UFH loops and the tank itself would need to be filled with antifreeze or suitable inhibitor that's good to at least -10 degrees.

 

Or just have a small "outside" circuit circulating heat from the ASHP to an input coil on the buffer tank, then the tank itself and the UFH only needs an inhibitor not antifreeze?

 

So that's the first question, what sort of tank?

 

Second question is control of the UFH circulation temperature.  Conventional wisdom is feed hotter water than you need to the manifolds and have a blending valve on the manifold to set the flow temperature.  But the blending valves seem expensive.  How about a dumb UFH manifold, and simply set the temperature in the buffer tank (with a tank stat) to the desired UFH temperature?  Only downside is it would not allow upstairs and downstairs UFH to run at different temperatures.

 

Third question, how large should the buffer tank be?

 

And last question, this will be a sealed system with inhibitor, so is it still a legionnaires risk so does it need periodic heating to a higher temperature or can it run forever at 30 degrees or whatever is required?

[Jeremy Harris]

I used an indirect buffer tank by accident - I ordered a direct one but Newark made an error and sent me an indirect one.  the upside of that is that the volume of (expensive) antifreeze/inhibitor is a lot lower.  If doing things again I think I'd fit a small thermal store as a buffer, with a mains pressure coil for DHW preheat, and take the hit on the cost of the extra antifreeze/inhibitor.

 

There's no legionella concern as the system is sealed and isolated from any water supplied to the house.

 

A TMV for the UFH isn't expensive and gives the ability to regulate the UFH temperature independently of the buffer, and that's very useful if you're thinking of using the buffer to preheat DHW.

 

I've found that a 70 litre buffer seems to be OK, it eliminates ASHP short cycling when heating and doesn't take up too much space.  Ours is fitted in the bottom of an airing cupboard, without taking up too much room, as it's about worktop height.

[ProDave]

Re the TMV, that typically adds £150 for a pump and mixing valve to each manifold. So that's £300 extra for upstairs and downstairs.  .  If I go the "tank regulates the temperature"

route all I need is the manifolds, which are a LOT cheaper than the blending valve and pump set, and a simple single circulating pump. Remember I am on an almost non existent budget, so £300 extra spend needs a very good reason.

 

I am keeping HW and heating completely separate so not bothered about using the heating buffer as a pre heat for DHW, that probably would not work ver well as they are some distance apart. There will be an unvented tank for DHW which will get pre heated by the heat pump on a separate cycle and topped up with an immersion.

[Jeremy Harris]

I'm pretty sure you need the UFH pump, so the TMV isn't really adding much.  The UFH manifold probably isn't going to work that well without a pump station to circulate water around.  Looking at our manifold system I can't easily see how there would be any significant flow around the UFH without a pump station, TBH, as the ASHP pump would just shunt water across the return side. 

 

Maybe you could try and connect the flow side of the manifold to the ASHP flow, but I'm not 100% sure that would work OK unless you could separately control the ASHP pump.  Here's a photo of our manifold, with the flow from the ASHP coming in at the bottom left, the upper manifold is the flow manifold the lower is the return and the ASHP return comes off at the lower right, after the thermally actuated valve:

 

 

[ProDave]

I am visualising a system with two pumps. One that will circulate heat from the ASHP to the buffer tank, controlled by the tank thermostat *. The second that will pump from the buffer tank to both mainifolds (each having their own 2 port valve, so only a manifold calling for heat will get water pumped to it).  That second pump should circulate heat through the UFH loops in a simple "no blending valve" setup.

 

As well as saving the cost of the manifold pump / blending valve kits for the two manifolds, it also avoids the need for an automatic bypass valve as there will never be the situation of the pump pumping with nowhere to go (which you get then the manifold is up to temperature and the blending valve has shut off)

 

* To give the full picture of what's in my head, the ASHP pump will have a 3 port valve so it will either deliver low temperature water to the buffer tank, or higher temperature water to the DHW tank, but never be used in the mid position supplying both together.

[PeterW]

4 minutes ago, ProDave said:

* To give the full picture of what's in my head, the ASHP pump will have a 3 port valve so it will either deliver low temperature water to the buffer tank, or higher temperature water to the DHW tank, but never be used in the mid position supplying both together.

 

So that's a standard W plan with a diverter valve - if you use a buffer tank then it makes life easy. 

 

This is as mine is planned - buffer is a 900mm direct copper tank from Newark Cylinders in their sale.

 

 

[ProDave]

So Peter, are you having blending valves on your UFH or doing as I am suggesting and just setting the buffer tank temperature?

 

I looked at cheap "standard" indirect hot water tanks, but you can only draw the hot water out right at the top, with the return going back in right at the bottom. Most proper buffer tanks have the inlet and outlet part way up, but you suddenly find you have to pay a LOT more for a "buffer" tank than you do a plain ordinary hot water tank. They are much the same thing except the position of the tappings.

 

How would a standard cheap indirect HW tank actually work as a buffer tank? my guess is not very well at all?

[dpmiller]

Get some Essex flanges and put the taps wherever you like.

[Jeremy Harris]

Newark were happy to fit the connections wherever I wanted, the price seemed to be the same for any layout I think.  A 70 litre tank wasn't expensive, I have a feeling that the shipping cost was a major element of it.

[PeterW]

28 minutes ago, JSHarris said:

Newark were happy to fit the connections wherever I wanted, the price seemed to be the same for any layout I think.  A 70 litre tank wasn't expensive, I have a feeling that the shipping cost was a major element of it.

 

Yep pretty much my experience too. 

 

The tank is a standard 900x300 and whilst having the tappings as you say, the basic 70-80 litres at 37c is here nor there when it comes to being used as a buffer as you won't get much stratification anyway. 

 

I have a Wunda mix valve and manifold - will be swapping it for an Esbe unit with a lower blending temperature as plan to run the floor at 31c. 

 

The idea of the buffer is to give the floor something to work with when the ASHP is in DHW mode - it also allows the ASHP to run for 15 mins which is what it should take to bring it up from 20c to 37c if I've worked it out correctly .... 

 

1 hour ago, ProDave said:

How would a standard cheap indirect HW tank actually work as a buffer tank? my guess is not very well at all?

 

It will be fine - better than the short cycling you will get on an ASHP as it struggles to modulate down to the 0.5 - 1kw you will need when you're at the end of the heat cycle. Also means you can prioritise DHW and not be cold ....

[ProDave]

2 hours ago, JSHarris said:

Newark were happy to fit the connections wherever I wanted, the price seemed to be the same for any layout I think.  A 70 litre tank wasn't expensive, I have a feeling that the shipping cost was a major element of it.

Out of interest how much did your Newark tank cost. Them and Telford are ones I will be looking for quotes from.  Can you save the delivery cost by ordering through a builders merchant, e.g. TP?

 

I'm in no hurry so also looking at ebay.

[PeterW]

It was in the sale and I needed a slim tank as it fits in the corner so I had it with 25mm of insulation on it - yes I'll get higher losses but I may box it in and give two sides some more when it's fitted properly. 

 

 

It cost me £100 - look in the clearance section on the site. 

[bassanclan]

If you have a system with a thermal store, ufh etc it's all filled with antifreeze and inhibitor, however if you introduce a tmv, then surely you are introducing clean cold water into the system and diluting the chemicals by an unknown amount. Or is it not really cold, but simply cooler water from the bottom of the thermal store you are introducing?

Edited July 19, 2017 by bassanclan

[PeterW]

4 hours ago, bassanclan said:

If you have a system with a thermal store, ufh etc it's all filled with antifreeze and inhibitor, however if you introduce a tmv, then surely you are introducing clean cold water into the system and diluting the chemicals by an unknown amount. Or is it not really cold, but simply cooler water from the bottom of the thermal store you are introducing?

 

The TMV is closed loop mixing of water returning from the UFH loops and then returning the remainder to the TS. Hot is blended in from the top of the TS / Buffer /ASHP as required. 

[bassanclan]

14 minutes ago, PeterW said:

 

The TMV is closed loop mixing of water returning from the UFH loops and then returning the remainder to the TS. Hot is blended in from the top of the TS / Buffer /ASHP as required. 

I had a thought about that and realised I was talking rubbish! It was too early in the morning for my brain

[Nickfromwales]

22mm TMV

 

Sorry to be a bit late to this one, but if using UFH then you need s TMV as a failsafe in case your thermostat goes wonky and you get full range heat going to the UFH. That can be dangerous, and cause a lot of damage to flooring.

If using the buffer for DHW preheat, I would not recommend going for a TS with an internal DHW coil. I know @JSHarris says he would go that route if doing this over again, but I had a very long chat with the tech guys at Telford who said that with a low temp ( for good CoP ) the yield from an integral DHW coil drops off quite significantly. They went on to say that the coils are rated to provide good rates of heat exchange when the temp is above ~55oC and at its best at >65-70oC. 

When I asked what size TS and coil would give eveninical yield at low temp they suggested no less than a 300L TS with the biggest ( 46kw iirc ) coil that will fit into  it. Only then is it really worth doing preheat that way. 

They went on to say using PHE's as @JSHarris has now would mean a much smaller TS / buffer could be used as they are far better at extracting heat from a low temp source than a coil would ever be. 

The killer is the drop in set / target temp, but another thing to remember is you can have pv divert to this TS as well as your UVC so maybe better to go bigger with less losses, and you'll need the TMVs to cap the output / flow temp according to how hot the immersion gets it. . 

[Jeremy Harris]

9 minutes ago, Nickfromwales said:

If using the buffer for DHW preheat, I would not recommend going for a TS with an internal DHW coil. I know @JSHarris says he would go that route if doing this over again, but I had a very long chat with the tech guys at Telford who said that with a low temp ( for good CoP ) the yield from an integral DHW coil drops off quite significantly. They went on to say that the coils are rated to provide good rates of heat exchange when the temp is above ~55oC and at its best at >65-70oC. 

When I asked what size TS and coil would give eveninical yield at low temp they suggested no less than a 300L TS with the biggest ( 46kw iirc ) coil that will fit into  it. Only then is it really worth doing preheat that way. 

They went on to say using PHE's as @JSHarris has now would mean a much smaller TS / buffer could be used as they are far better at extracting heat from a low temp source than a coil would ever be. 

The killer is the drop in set / target temp, but another thing to remember is you can have pv divert to this TS as well as your UVC so maybe better to go bigger with less losses, and you'll need the TMVs to cap the output / flow temp according to how hot the immersion gets it. . 

 

Nick,

 

This is what I wrote:

 

On 16/07/2017 at 15:19, JSHarris said:

I used an indirect buffer tank by accident - I ordered a direct one but Newark made an error and sent me an indirect one.  the upside of that is that the volume of (expensive) antifreeze/inhibitor is a lot lower.  If doing things again I think I'd fit a small thermal store as a buffer, with a mains pressure coil for DHW preheat, and take the hit on the cost of the extra antifreeze/inhibitor.

 

 

I'd not recommend using an indirect small buffer as we have, but having said that I did calculate the heat transfer rate for the lower temperature and it massively exceeded the ASHP output.  Newark were happy to replace the wrong tank they'd sent, and gave me the surface area of the coil in it, so I could see whether it had any impact.  As it turned out, the indirect coil was rated for around 40kW at 55 deg C, so handled 6kW at 40 deg C with no problem at all.  In practice we get no measurable difference between the buffer temperature and the ASHP flow temperature at all, and the heat up rate from cold seems to match the ASHP output.

 

As above, I'd not go down this route by choice, but I made do with what we received by accident.  The reason I suggested a small direct TS instead of a buffer was just to eliminate the need for a PHE if using it for DHW preheat.  It would be a lot simpler to just run the cold water into a small TS coil for preheat and then run it to the Sunamp PV, electric boost heater or whatever.

[Nickfromwales]

8 minutes ago, JSHarris said:

  If doing things again I think I'd fit a small thermal store as a buffer, with a mains pressure coil for DHW preheat,

Yes that's the snippet I'm referring to. Just the tech guys were convinced, and could back up the fact that your earlier proposed setup would be the better option purely based on the remit of extracting heat energy in reasonable quantities, e.g. when DHW is being drawn at high volume. 

Telford weren't convinced about the yield from the smaller TS / DHW coil, as I asked them about an install for a household with reasonable / high rates of DHW consumption. It fear your DHW consumption is far less than the figures I was proposing and wonder how your existing setup, facts now in order thank you , would cope with a couple of teenage girls in the mix. 

The angle I approached Telford with was "TS plus coil or TS plus PHE" with which they replied with the arguments aforementioned, eg a coil would need to be much bigger, and in a bigger ( min 300L ) TS to be of practical / economical benefit. 

Just to clarify the coil side of things, do you preheat via a PHE now?

[Jeremy Harris]

Yes, preheat is via a "44kW" rated PHE, but that's pretty much the same rating as the coil in the indirect buffer tank.  The rating is very highly dependent on the actual temperature differential - halve the temperature differential and you halve the heat transfer rating.

 

What I was thinking of, as a good solution for  combined ASHP buffer tank and DHW preheat system, was a direct TS of around 70 litres or so, and with a standard DHW coil in the top section.  There's no point in having more than about 100 litres, as the buffer only has to provide preheat to the DHW, so the power needed is around half that needed if the buffer/TS was running at a higher temperature.  The ASHP will provide constant top up to the TS when it cools a bit, too, acting as an additional "instant heat" source, in effect.

 

I would hazard a guess that the Telford guys were thinking in terms of "normal" DHW temperature, rather than just preheating to maybe 30 deg C or so.  It makes a big difference just dropping the temperature requirement a bit.  This only makes sense if you're running an ASHP, want to keep the COP high and have another means of boosting the water temperature up to normal DHW temperature.  With a single Sunamp PV, the effective combined capacity with a 70 litre preheat buffer plus a 6kW ASHP is close to double the capacity of the Sunamp PV on its own, plus the buffer/TS could still do it's normal job of providing a direct buffer store for the ASHP when it's running the UFH.

 

[PeterW]

So......

 

Youve now got me thinking about using my baby buffer tank with an external PHE....! I have a direct tank with no coil, so would need PHE plus flow switch and pump to drive the transfer however it would be a pretty simple thing to fix... Only issue is that when the ASHP is running I need it to not divert through the PHE (although I'm assuming resistance through the PHE would be higher than through the buffer tank) as I don't really want a zone valve in the way.

[Jeremy Harris]

I used a 44kW PHE plus a small Wilo pump and a solid state 3/4" BSP flow switch in the pre-heated DHW pipe to turn the pump on and off.  It thermosyphons well, so the PHE is always at the buffer tank temperature, which means that pre-heat is available straight away, it doesn't needs to wait for the slug of warm water from the buffer to be pumped around.

[PeterW]

Got any supplier links @JSHarris...?

[Nickfromwales]

The guys at Telford knew that it was low temp preheat and I told them a hysteresis of 30-35oC was to be assumed. 

In the words of Forest Gump, "im not a smart man", but my gut instinct told me they were over cautious in their response. Most big jobs I've done ( 14 bed student digs, with 9 rooms ensuite and 3 shared full bathrooms ) I've designed myself from my 'sixth sense'. They've all worked either well, very well or too bloody well, but never been a fail, so when I designed a pre-heated system for use in a whopper of a house I decided on a 500L TS for pre heat via a coil, and another 500ltr UVC for DHW. When questioning myself about the size of the preheat tank, as in can it be any smaller, I contacted Telford. 

@JSHarris. How do you think your setup would cope with a family of 5 instead of you and SWMBO, if you don't mind me getting some feedback on the hypothetical ? Ta. 

[Jeremy Harris]

28 minutes ago, PeterW said:

Got any supplier links @JSHarris...?

 

The PHE was this one from Germany: http://shop.wiltec.info/product_info.php/language/EN/info/p3855_-changeur-de-chaleur-thermique-inox-20-plaques-max-44kW.html

 

The flow switch was a Gentech FS-02: http://uk.farnell.com/gentech-international/fs-02/flow-switch-noryl-dc-1l-min-10bar/dp/1006767

data sheet here: http://www.farnell.com/datasheets/1702844.pdf

 

The pump was a low power Wilo, this one: https://www.pumpsukltd.com/yonos-pico-15-1-4.html but any small circulating pump would work OK, as the flow requirement is pretty small and the is very little head to work against.

[Jeremy Harris]

37 minutes ago, Nickfromwales said:

The guys at Telford knew that it was low temp preheat and I told them a hysteresis of 30-35oC was to be assumed. 

In the words of Forest Gump, "im not a smart man", but my gut instinct told me they were over cautious in their response. Most big jobs I've done ( 14 bed student digs, with 9 rooms ensuite and 3 shared full bathrooms ) I've designed myself from my 'sixth sense'. They've all worked either well, very well or too bloody well, but never been a fail, so when I designed a pre-heated system for use in a whopper of a house I decided on a 500L TS for pre heat via a coil, and another 500ltr UVC for DHW. When questioning myself about the size of the preheat tank, as in can it be any smaller, I contacted Telford. 

@JSHarris. How do you think your setup would cope with a family of 5 instead of you and SWMBO, if you don't mind me getting some feedback on the hypothetical ? Ta. 

 

It would need to be beefed up, Nick.  It can deliver a couple of showers, one after the other, without a recharge, but then the Sunamp PV will be out of capacity, and there would be a recharge time of around 30 minutes for the buffer, with DHW being reliant on the electric instant water heater.  Youd really need two Sunamp PVs, or a Sunamp Stack, for a bigger system.  The buffer tank could be increased in size, but the smaller one has the advantage of a faster recharge time from the ASHP, so it depends on how big a gap there is between large DHW demands.