heating/cooling application

[amb5l]

I'm working on controlling (from Loxone, via Modbus) a Daikin VRV IV air source heat pump. This is a high temp / low temp split system that can generate hot water, drive underfloor heating, and also blow cooled or heated air from fan coil units in certain rooms (they can warm up a cold room very much more quickly than via underfloor heating).

Would be grateful for any advice on the best way to efficiently control all this, especially on the following points...

1) The Daikin system can't generate high and low flow temperatures at the same time. It either provides a high temp flow for hot water, or a low temp flow for underfloor heating. When there is demand for both, it alternates, changing over every 30-45 minutes. Note that there is no direct visibility of what it has chosen to do, but you can infer it from flow temperature registers (or by looking at the motorised valve!) So if Loxone asks the heating, it may or may not get an immediate response, and that response might be diluted by the hot water workload. I worry that this will confuse an Intelligent Room Controller's learning process.

2) I'm not clear about how an IRC makes use of a second/supplementary heating source (zone). Is it just a booster for high demand scenarios? Does anyone understand this aspect clearly?

3) The fan coil units are a little noisy, even at low fan speeds. I'd like to stop them running at night. But again I might confuse the IRC if I interfere in it's control of a heat source (by suppressing it at night).

4) I haven't looked at the Intelligent Temperature Controller. Would it help? Would an external temperature reading be useful?

Thanks in advance...

[Simon Still]

On Monday, 30 October 2017 20:55:11 UTC, amb5l wrote:

I'm working on controlling (from Loxone, via Modbus) a Daikin VRV IV air source heat pump. This is a high temp / low temp split system that can generate hot water, drive underfloor heating, and also blow cooled or heated air from fan coil units in certain rooms (they can warm up a cold room very much more quickly than via underfloor heating).

What sort of building is it controlling? 

What's the level of insulation?

How often is it occupied? 

Is the system installed yet or do you still have options on how you can configure sensors/UFH zones etc? 

 

[amb5l]

It's a large 5 bedroom home, new build and fairly well insulated by modern standards - but not passive.

The configuration is "upside down", living space upstairs, bedrooms downstairs. A/C specified because leaving a bedroom window open on a warm night is not good for security / peace of mind (don't think the architect considered that). Family of four in residence, one mainly at home, the rest out at work during the week.

All installed and being brought up / refined gradually. Considerable flexibility on sensors and zones. There are around 35 Loxone Tree valves spread across 4 manifolds for the UFH. At least one temperature sensor per room/zone.

Upstairs is very open plan, roughly divided in two but with a large (15 sq mtr) opening connecting the spaces. I am therefore assuming it's best to treat it as a single area with a single IRC (?) rather than have one for each space.

[Andrew B]

My system is along the same lines.  I have a large thermal storage so that my heat pump can work primarily during the day when it is more efficient, and a dual loop system.  The primary loop is heated from the heat pump / thermal store and is at the maximum temperature possible at any given time.  A secondary loop feeds the radiant loops in the ceilings (same principle as underfloor).  There is a pump that "injects" heat from the primary to the secondary loop (and a return line from the secondary to the primary, obviously).  I operate this injection pump under a thermostat control based on the temperature I want running through the radiant pipes.  The primary loop feeds a heat exchanger tied into the ventilation system, and there is also a DHW pre-heat loop that comes before the on-demand hot water heater.

[Simon Still]

On Tuesday, 31 October 2017 18:27:12 UTC, amb5l wrote:

It's a large 5 bedroom home, new build and fairly well insulated by modern standards - but not passive.

Ok - there's a long way between building regs and passive on insulation and airtightness but from what you've said I'm assuming you're beyond regs.   

We're 2 years into a living in our near-passive self build. Before that we were in a new build well insulated flat, before that a draughty victorian end of terrace.  My view is the better insulated the house the less complex the heating system control needs to be. If you've got very high heat loss it makes sense to heat rooms only when you're in them and you then need to heat them up quickly. In a well insulated house temperatures are much more stable and you don't really need fine control but you do need to avoid overheating (because once a room is too warm it will stay too warm).  In winter with the heating off for a week (when we went on holiday) the internal temperature only dropped to abut 17C 

We've UFH in concrete on the two lower floors and under wood floor (in an overlay system) on the bedroom floors).  The UFH runs continuously (i think 5am to 11pm) with the flow temperature varying based on the outside temperature using a heating curve (It sounds like you can do this with the ITC).  The flow temperature is normally only slightly above room temperature.  Even when below 0C I think the concrete is still only at about 24C.  The advantage of this is that you don't overheat - when the room is at temperature the delta between air and floor temperature is negligible and no more heat transfer happens.  From my understanding your heat pump will work more efficiently if it's only raising the water temperature slightly. 

Our house is treated just as two zones - based on the floor type as the flow temperature needs to be slightly higher under wood. The manifolds just have manual valves on them. There are no thermostats in the rooms - it works purely off outside temperature.  This needed a bit of trial and error over the first winter monitoring the internal temperatures with a min max thermometer and tweaking the heating curve but last winter I just turned it on in October and off around March.  

I've actually found the bedrooms don't need any heating and turned those circuits off manually and they run just slightly cooler than the living rooms.  

This approach would probably work for you.  I don't think you need to worry about the hot water heating - there will be enough latency in your floor heating that periods when the unit is heating water won't affect room heating.  I suspect you'll never need hot air heating

I don't know much about cooling but you could have a simple switch in the interface to move it between heating and cooling modes (or control it with a calendar).  For Air cooling my guess is you'd just want a simple IRC or ITC block.

[Rob_in]

I'm pretty interested in this because we have a smaller house into which the Daikin Altherma heat pump has just been installed. It's yet to be commissioned (awaiting definitive power connection) but when it is I want to monitor and tweak it with Loxone. When we chose this months ago I was told the Modbus interface for this model (CB series) was 'coming soon' and notice that's still the case. However, there is now an ethernet box you can connect to this but Daikin are reluctant to share any interface details (which is odd given how freely available Modbus info is) so you can only control with a smartphone app. If the Modbus interface doesn't appear soon I could reverse engineer something on the ethernet one, but prefer not to.

Example usage: there is no heating running in the house right now, but when the sun shines it can easily reach mid 20s inside because of south facing windows (don't worry - the eves are designed so no sun enters the house during summer!). At around 10am the other morning it was 11 degrees outside but still 22 inside due to solar heat gained from the day before.

Because of this I want to use weather forecast and/or sunshine sensor to disable the UFH, or at least reduce the curve controlling it when we know there will be free solar heat in winter. Conversely, as UFH is slow to respond would like to turn it on if the forecast says an unseasonable cold snap is coming (well, maybe!).

I haven't bothered to spec anything to control the UFH manifolds because it's a small house with a heat exchange ventilation system so figure temperatures can be kept similar all over just relying on the heat curve and not messing with flows to the rooms.

Just keen to hear how similar control is going for others. Oh, and how the interface is made, Modbus, ethernet thingo or something else.

Cheers,

Robin

[Simon Still]

On Wednesday, 1 November 2017 19:48:30 UTC, Rob_in wrote:

Example usage: there is no heating running in the house right now, but when the sun shines it can easily reach mid 20s inside because of south facing windows (don't worry - the eves are designed so no sun enters the house during summer!). At around 10am the other morning it was 11 degrees outside but still 22 inside due to solar heat gained from the day before.

Because of this I want to use weather forecast and/or sunshine sensor to disable the UFH, or at least reduce the curve controlling it when we know there will be free solar heat in winter. Conversely, as UFH is slow to respond would like to turn it on if the forecast says an unseasonable cold snap is coming (well, maybe!).

See my suggestion above.  If you run the floor constantly at a very low temperature you don't need to worry about solar gain.  As the room approaches floor temperature the delta between the two is small and heat transfer is negligible - it's self regulating.  This morning it's 7C outside the the floors are running 24C and 22C.  At 11C outside they'd be running at the minimum temperature (i think they're set to bottom out at 20C).  

[Rob_in]

On Thursday, 2 November 2017 10:10:43 UTC+1, Simon Still wrote:

If you run the floor constantly at a very low temperature you don't need to worry about solar gain.  As the room approaches floor temperature the delta between the two is small and heat transfer is negligible - it's self regulating.

Sorry, I don't follow. Did I miss something?

I though UFH worked like this: In standard configuration the UFH boiler (in our case an air source heat pump) knows the outside temperature and uses a heat curve and the desired inside temperature to calculate how much heat needs pumping into the house (it doesn't use a room thermostat because UFH is so slow to respond it will overshoot and miss the desired temperature). On a cloudy day, all that heat comes from the UFH so this works fine. If it is sunny though (or we light the fire - small log burner mainly for ambiance), then the heat curve calculation needs to be adjusted to take the added solar/fire inputs.

You are saying (I think) just knock the curve used by the UFH boiler in calculations down depending on actual room temperature. The closer the actual measured temperature comes to desired temperature the lower it gets until the UFH boiler thinks (for example) to maintain 22 degrees inside, even when it's 7 degrees outside, no power is required?

It is actually possible to order a room thermostat for our boiler but didn't order one. Is the above how that would work?

Cheers :)
Robin

[Simon Still]

On Saturday, 4 November 2017 07:38:19 UTC, Rob_in wrote:

On Thursday, 2 November 2017 10:10:43 UTC+1, Simon Still wrote:

If you run the floor constantly at a very low temperature you don't need to worry about solar gain.  As the room approaches floor temperature the delta between the two is small and heat transfer is negligible - it's self regulating.

Sorry, I don't follow. Did I miss something?

I though UFH worked like this: In standard configuration the UFH boiler (in our case an air source heat pump) knows the outside temperature and uses a heat curve and the desired inside temperature to calculate how much heat needs pumping into the house (it doesn't use a room thermostat because UFH is so slow to respond it will overshoot and miss the desired temperature).

That's a good standard config. A lot of UK UFH doesn't get set up like that - people try to control it like a radiator and use internal room thermostats and short heating periods: "i only want to heat the house when I'm home so the heating should only be on for a few hours in the morning and evening".  To get it to work (badly) like that you need to run the floor at a fairly high temperature when it is on and have it switch on some hours before you want the room to be at desired temperature.  The UFH flow temperature is fixed (most older UK gas boilers have fixed output and manual, fixed, temp mixers on the UFH). Unless I'm misunderstanding that's what you'll get with a Loxone Intelligent Room Controller and a room stat (i don't see how Loxone could 'learn' if flow temperature was varying but maybe it's more advanced than I appreciate).

On a cloudy day, all that heat comes from the UFH so this works fine. If it is sunny though (or we light the fire - small log burner mainly for ambiance), then the heat curve calculation needs to be adjusted to take the added solar/fire inputs.

You are saying (I think) just knock the curve used by the UFH boiler in calculations down depending on actual room temperature. The closer the actual measured temperature comes to desired temperature the lower it gets until the UFH boiler thinks (for example) to maintain 22 degrees inside, even when it's 7 degrees outside, no power is required?

Not quite.  I'm saying that if your house is well insulated enough and you run your UFH near constantly (ie maybe 18 hours a day) you get to a floor temperature low enough that you don't need to worry about the extra heat inputs.  If the floor surface temperature is the same (or very close) as the room temperature then it can't heat the room any more so a few hours of solar gain, or input from a fire, doesn't matter.  

https://www.underfloorheatingsystems.co.uk/self-install-information/questions-and-answers/what-is-the-water-flow-temperature/ for example talks about flow temperatures of 35C-60C. I don't think either of my floors get above 30C even when it's some way below 0C outside.  I specced close pipe spacing as well - a 'budget' UFH with less pipe probably needs higher flow temp to get the same floor temp.

[Simon Still]

On Saturday, 4 November 2017 07:38:19 UTC, Rob_in wrote:

. If it is sunny though (or we light the fire - small log burner mainly for ambiance), then the heat curve calculation needs to be adjusted to take the added solar/fire inputs.

With higher flow temperature I suppose you could easily add a room stat to override the heating curve.  I know it was an option we could add for our system and we put the cable in but havent needed it.  

The other experience I have is with my Uncle.  Highly insulated house with wood pellet burner UFH - that had been configured with radiator style heating periods and just wasn't working.  Reconfigured to run longer it's much better.  

However, despite it being an airtight, near passive house, he insisted on a large open fire (chimney was blocked for the air-tightness test) so you sit there in winter sweating as big logs roar away every night.  So much uncontrolled heat going into the room that the UFH temperature/control is irrelevant. 

[Rob_in]

On Saturday, 4 November 2017 12:27:10 UTC+1, Simon Still wrote:

https://www.underfloorheatingsystems.co.uk/self-install-information/questions-and-answers/what-is-the-water-flow-temperature/ for example talks about flow temperatures of 35C-60C. I don't think either of my floors get above 30C even when it's some way below 0C outside.  I specced close pipe spacing as well - a 'budget' UFH with less pipe probably needs higher flow temp to get the same floor temp.

35C-60C? LOL... your example of "the floor surface temperature is the same (or very close) as the room temperature" is what I had in mind. I was going to configure the boiler to put 28C max into the floor and thought that would even be overkill!

Regarding that, "...it can't heat the room any more so a few hours of solar gain, or input from a fire, doesn't matter." I think you mean that if the boiler is sending water out into the UFH loop at 24C say, and it comes back at 24C, or even above that (for example, because the sun or fire is heating the floor) then the boiler won't bother putting any extra heat into the loop and the 'heat pump' part of the system will be shut off automatically (the system will keep circulating water to even out temperatures around the house though).

That sounds fair enough. I guess we will wait and see how much actual input from sun through the windows we get in winter. Too much free energy can't be a bad problem to have ;)

Robin

[amb5l]

Really useful and interesting feedback so far, thanks all.

The Daikin system Modbus interface is handled by a Realtime Controls RTD-W which drives the hydrobox, and 6 RTD-NETs (one per fan coil).

The stuff about using low flow temperatures and (almost) leaving everything on is really interesting and makes sense to me. I am using a high flow temp at the moment (45degC) with towel radiators in mind, but could lower the UFH flow temp while still giving the towel rads a burst of heat while the DHW is being heated up (overnight and late afternoon). And I will study the ITC; I can control flow temp via Modbus.

The house has large amounts of glass - many floor to ceiling windows, with many of these facing south. A bit of late autumn sun makes an amazing difference in the living space. And there is a fireplace. So again, a low floor temp to self-limit heating /  stop overheating really makes sense. Although there may be some customer complaints, perceptible warmth from bathroom floors is very popular!

On Monday, 30 October 2017 20:55:11 UTC, amb5l wrote:

[Simon Still]

Do you have a ventilation system? With our Heat Recovery Ventilation towels dry quickly without needing any towel radiators.

With the UFH, even if it’s just running at 21C (which i think is the minimum circulation temperature) the floors don’t feel cold.

[Rob_in]

On Saturday, 4 November 2017 12:27:10 UTC+1, Simon Still wrote:

...a few hours of solar gain, or input from a fire, doesn't matter.

Well... it was -4C overnight here and a max of just 10C during the day. There is no heating in our house (still waiting connection), and it was nippy inside this morning due to a few cloudy days cooling down, but this was the temperature in the cellier, which is a small room off the lounge (the lounge having large south and west facing windows & I haven't installed the sensor there yet):

It was lovely and warm in the house in the afternoon so on a day like this doubt one would want the heating on much past lunchtime.

I'm just sceptical that without some kind of control linked to a weather forecast (that knows it will be sunny all day) that the house would overheat quite easily.

We'll find out soon enough I guess but keen to hear any more tips.

Robin

[Simon Still]

On Tuesday, 7 November 2017 21:36:03 UTC, Rob_in wrote:

It was lovely and warm in the house in the afternoon so on a day like this doubt one would want the heating on much past lunchtime.

Don't think about wanting the heating on or off - the heating is always 'on' during the heating season.  Water circulates in the UFH the whole time but if the return temperature is the same as the outbound then the heat source doesn't need to add any heat to it. If one room is getting a lot of solar gain then it's redistributing that heat to the other rooms.

It's 7C here now.  Rooms are at 19.4C, Concrete floors are running at 22.5C, Wooden floors at 23.6C.  

[Sergey Rozenblat]

Unless you have huge windows. My living room (double high ceiling with full height window and black floors) overheats to 28-29C (around 0C outside). Of course UFH switches off as soon as sun appears, but this is not enough in winter when sun is low. Sometimes temp is still above set until next morning so UFH may be off for days. So in my case I think only window blinds will save me.

[Duncan]

ive built in electric blinds with 90% light obstruction and electric roof windows to integrate into loxones 'overheat' output of the IRC and together get an upper limit of around 25-26 degrees on a summer day in the south/west facing areas - seems to work quite effectively at limiting overheating.