Heat Pumps and Zones (or not?)

[SimonH2]

I can see the benefits of using weather compensation, I understand how it would work regarding flow temperatures, but how this works in practice with keeping a space heated without using room thermostats needs more investigation on my part.

Doesn't need much research - it wouldn't work !
It can help make a system work better (by varying rad temps with outside temp, it can minimise the problems of 'all or nothing' hot rads in mild weather and not enough heat in cold weather), but without internal sensors/controls it can't work (except in certain very controlled and consistent settings that don't include domestic properties) to effectively control room temperatures.

Example:
Our lounge is upstairs and has large windows and a high vaulted ceiling. When the sun comes out it's like a greenhouse and doesn't need heat - even if it's cold outside. But if the sun isn't out, it can need heat even in warm weather.
And for good measure, because of the menagerie, the dining room door at the bottom of the stairs is normally open, meaning the the dining room heating effectively heats the lounge - except when it doesn't because we've shut the door to be cosy downstairs.
And that's just the one room with multiple disturbances.

The dining room is semi-open with the kitchen. So the kitchen can be a heat sink sucking heat away, or it can (when cooking for a big meal) put so much heat across that we need to open the window. And when the door is open, it can stay cold as the heat buggers off upstairs.

Our spare bedroom is also my office most of the time. When I'm WFH it needs heating during the day. When it's a bedroom (the infrequent occasions we have guests) it needs a little heat early morning and evening. When neither applies it just needs enough heat to avoid damp problems.

So if anyone suggests weather comp can control that, I can challange them to explain why they think the laws of physics don't apply

I strongly suspect the simplest way of managing multiple rooms/zones is :
Each room/zone has an internal sensor/controller that sets a flow temperature demand. This may include an element of weather comp.
The heat pump then provides the highest of the demanded flow temperatures. This will be more than some room/zones need so they will be shut off via rad/zone valves when they reach set temperature.
So the heat pump should run continuously until it either bottoms out at min output power, or all zones/rooms are satisfied - in either case it means cycling unless you have the ability to dump some excess geat into (e.g.) a DHW cylinder.

In our house I've installed a thermal store (heated by gas boiler) - SWMBO likes the idea of a wood burner, we've room for solar panels when I have the spare cash, and of course there's scope for adding a heat pump. The CH loop runs off this with modulating pump and mixer valve. My aim is a control setup as I've described where the CH flow temp is controlled to suit the highest requirement, and other rooms turn on/off as required - with wall mounted 'stat' in each room.
The store temp can then be controlled so as to meet the higher of CH or DHW temps (there's multiple sensors to suit). If heat pump added, the control regime could be to keep adding heat to the store until COP drops, then turn off for a longer period until the store temp drops again
It's all going to need custom controls (I doubt there's anything on the market at an affordabke price), and I'm struggling to find time to sit down and do some Arduino programming

 

[Ivor Windybottom]

I've seen timer/thermostats that use Modbus to communicate with gas boilers, so can control the boiler's flow temperature. One of these in your hallway could control a gas boiler so it gradually tails off its power as it approaches the setpoint, avoiding overshooting and the on/off swings of temperature.

However, I haven't looked any more than a quick glance but don't think that most heat pumps will work with modbus. But even if they did then you probably wouldn't want to, as it works most efficiently if stays within a known range of the outside temperature. If a controller commanded it to produce a really hot temperature while it was subzero outside then it could use massive amounts of power trying to achieve the impossible.

I've no idea whether a cluster of modbus controllers would be able to chat between themselves and vote on the best temperature based on their separate differing requirements. While undoubtedly possible, I doubt that anyone's made a domestic thermostat that can do this.

My conclusion is that it's best to leave the heat pump to decide its own flow temperature based on the outside temperature. Then use zone valves and TRVs to control the interior temperature by blocking (zone valves) or throttling (TRVs) the flow of water.

The concept of leaving all the valves open just seems idiotic, it's back to the 1960s really. The heat pump evangelists seem to be chucking away several decades of knowledge in pursuit of willy waggling their COP numbers on youtube, while wasting power having room temperatures that are needlessly hot.

 

[SimonH2]

I've seen timer/thermostats that use Modbus to communicate with gas boilers, so can control the boiler's flow temperature. One of these in your hallway could control a gas boiler so it gradually tails off its power as it approaches the setpoint, avoiding overshooting and the on/off swings of temperature.

However, I haven't looked any more than a quick glance but don't think that most heat pumps will work with modbus. But even if they did then you probably wouldn't want to, as it works most efficiently if stays within a known range of the outside temperature. If a controller commanded it to produce a really hot temperature while it was subzero outside then it could use massive amounts of power trying to achieve the impossible.

The thing is, if that's what's needed to heat the house, then that's what's needed. If the heat pump can't cope then something isn't sized right for the installation - or it's just one of those 0.1% of the time that the right sized system isn't design to handle.

My conclusion is that it's best to leave the heat pump to decide its own flow temperature based on the outside temperature. Then use zone valves and TRVs to control the interior temperature by blocking (zone valves) or throttling (TRVs) the flow of water.

That means, a lot of the time, a flow temperature higher than is needed. You'd need to set the weather comp to suit the worst case conditions in the house, meaning that for the other 99% of the time the flow temperature could be cooler. You could, for example (and assuming it's an available option), set the heat pump to not go below a certain CoP, meaning that if the internal controls ask for something too high, the heat pump will limit the flow temp and it'll take longer to heat a room.

The concept of leaving all the valves open just seems idiotic, it's back to the 1960s really. The heat pump evangelists seem to be chucking away several decades of knowledge in pursuit of willy waggling their COP numbers on youtube, while wasting power having room temperatures that are needlessly hot.

It does look a lot like that.

I've seen some really mind boggling ideas around.
One is that all your heaters must be in series, and room by room control is by turning the fans up/down. Firstly, who want a fan heater unless they truly don't have space for a similar capacity radiator. And who wants a system that is incapable to being turned fully off in a room where it's not needed.
We have some fan-coils in our church, and as installed their lowest speed was too noisy to leave them on during a service. Even when off they act (once the water is nice and hot) as a convection heater, which is handy. Plus I modified them with some resistors to allow an "extra low" fan speed which means they can stay on most of the time. But as originally installed, the system was "rubbish" - the fan coils sucked so much heat out of the system that the water stayed cool, then when switched off for the service the place went cold for half an hour until the pipes got hot enough to work on their own (old system designed to run at high temperatures so large cast iron pipes act as the radiators). That's another system crying out for a bit of Arduino programming

 

[MNW67]

I'm finding this thread really intriguing. So, I've done a quick, (very) broad brush stroke heat loss comparison based on the info you have posted. Nothing to do with what was in the Heat Geek presentation. I think doing things your way might reduce overall heat loss by around 20%. Please feel free to ignore or point out that it is nonsense; it has been fun so I will not be offended. I have assumed two equal size zones. Temperature falls are not linear. Two schedules a) and b). Based on typical outside average temperature of +3C during December to February. I hope I've not made any stupid mistakes.

Zone 1:

a) 22C from 4pm to midnight then heating off until 4pm the next day. During this time the temperature drops to 12C but the average is 15C.

b) 22C from 8am to midnight then heating off until 8am the next day. During this time the temperature drops to 15C but the average is 18C.

Zone 2:

a) 22C from 8am to 4pm then heating off until 8am the next day. During this time the temperature drops to 12C but the average is 15C

b) 22C from 8am to midnight then heating off until 8am the next day. During this time the temperature drops to 15C but the average is 18C.


Schedule a) has an average temperature across 24 hours of 17.33C and schedule b) has an average of 20.66C.

 

[Ivor Windybottom]

That's another system crying out for a bit of Arduino programming

Personally, I'm of the opinion that it's really not worth scratting around for that tiny percentage of extra efficiency. Much better to invest in a decent heat pump and big radiators so it can heat the place sufficiently without needing a high temperature.

Rads in series would be stupid. Only the first would be hot, the rest progressively colder. That's why they're plumbed in parallel and balanced using their outlet valves, to spread the flow across them.

I fully intended to get an air-to-air system. Then along came the £7500 grant for an air-to-water system, which is not available for air-to-air. But I was still pondering either way. Then we went to a small business that had a decent new air-to-air system installed. We sat in the waiting area under one and concluded that the noise, while definitely background level in a business, would be downright annoying in a home setting. For the benefit of cooling during the odd heatwave it's just not worth putting up with fans of any sort. The £7500 was the final nudge I needed towards radiators. Fan assisted radiators sound like the worst of all worlds - all the noise, but without the cooling ability.

I've seen before that when something new and mysterious comes along people cook up all sorts of mad ideas. Putting up a youtube video with a controversial title, e.g. don't do things the usual way, will get more views (and therefore revenue) than one saying that all conventional wisdom is correct. So there's a vested interest in suggesting the opposite of what many think.

I'm about to order my contenders for controlling it all. The requirements are reasonable price, hard-wired so no batteries and a phone app that integrates a number of them into a single interface. I think I've found a contender, I will buy them and bench-test them.

 

[Mike13]

Some of it seems counter intuitive. I ended up last night trying to work out the cost of heating fifty tonnes of masonry through 1.5C!

This takes you into a detour of looking at the "Decrement Delay" of the building:

Dynamic thermal properties

A free tool for calculating the thermal properties of construction elements.

www.concretecentre.com

 

[snb1]

one thing that many have forgotton about is how a heat pumps own control logic works. For many heat pumps when a demand is created for heating it will start and run at low output for a few mins then start to increase it's output as required.
If someone uses enough hot water then the systems will normally switch over to heat hot water ( unless time control prevents)
But on zoned systems ( especially with underfloor) when a zone is being heated and then another asks for heat the return temp drops quickly and in most cases the heat pump either defaults to minimum putput or goes into a waiting period. On underfloor you don't notice. But with radiator systems this is far more noticable. One of the ways to reduce this effect is to fit a buffer vessel ( 50lt will normally be more than big enough). This manages to solve the minimum flow issue with a heat pump and reduces the effect when zones open and close.

As long as the hat pump can produce enough heat when it is required you are ok, but to be a good heatloss by room calculation needs to be completed a long with the heat emmitters. over sizing heat emmitters by up to 30% allows the system to reduce the heat up time for zoned setups but does add cost. Depending upon the calculated heat losses per room against increased install costs would probably sway your decision. Obviously the gretaer the heat loss per room gives you more savings from being zoned, but that can come at a price of reduced comfort.

You always need to think though that the simpliest system that works for you is the best. ( more electronic control can create more fault finding at a later date)

 

[BlueLoo]

Women ruin all of this calculation and effort btw.
Especially hormonal women.

Go speak to any HVAC engineer. They will all regale you of tales of mental torture in trying to get a ambient room temp satisfactory to all. It's an impossible task.

I say this jokingly but it is a real issue, as is the secondary structural problems they present by demanding windows be opened at all times, radiator covers, vertical rads, shoving anything they can find to put Infront of them. Hanging clothes off them, etc.

Then, you also have the human body interface problem and the dumbing down of society which seems to demand plug-n-play automatic solutions for most things these days. My neighbour uses his thermostat as an on-off switch. explaining to him how it works is an annual ritual.

How to cope with all that input into a control system for a slow reacting heating system?

Best of luck.

 

[SimonH2]

I ended up last night trying to work out the cost of heating fifty tonnes of masonry through 1.5C!

I would suggest that it's not very important in terms of heating costs. After an initial warm up, it's just stored heat.
Where it does come in is when considering how much energy can be saved by backing off the heating in unused rooms. If you have massive thermal masses and good insulation, then there will be limited drop in temperature, and so little saved in heat losses. At tge other extreme, the temperature reduction and hence heat loss would be significant.

Personally, I'm of the opinion that it's really not worth scratting around for that tiny percentage of extra efficiency. Much better to invest in a decent heat pump and big radiators so it can heat the place sufficiently without needing a high temperature.

I'm not doing it just for efficiency - mostly it's to get the comfort and control I want. I've been progressively fitting larger rads as I work on rooms, generally just fitting the largest that fits without going OTT. The house had some ridiculously small rads when we bought it - including a complete waste of space designer rad that couldn't heat the lounge even with the boiler turned up to max (82C) !

Rads in series would be stupid. Only the first would be hot, the rest progressively colder. That's why they're plumbed in parallel and balanced using their outlet valves, to spread the flow across them.

This was specifically with fan-coils - so control by fan is possible. But I'd not fit them in a house unless it was the only option.

Then we went to a small business that had a decent new air-to-air system installed. We sat in the waiting area under one and concluded that the noise, while definitely background level in a business, would be downright annoying in a home setting.

I agree. It's not just the fan noise either, there's an unavoidable noise from the refrigerant boiling in the evaporator.
Another downside is control of multiple zones. For a single zone, inverter drive of the compressor can vary the off-coil air temperature. For multi-zone, you either have on-off control on each indoor unit, or multiple outdoor units. On-off control in an air system is both annoying (changes in noise), and results in rapid temperature changes.

Fan assisted radiators sound like the worst of all worlds - all the noise, but without the cooling ability.

Yes

I've seen before that when something new and mysterious comes along people cook up all sorts of mad ideas. Putting up a youtube video with a controversial title, e.g. don't do things the usual way, will get more views (and therefore revenue) than one saying that all conventional wisdom is correct. So there's a vested interest in suggesting the opposite of what many think.

Agreed

 

[SimonH2]

Women ruin all of this calculation and effort btw.
Especially hormonal women.

Stand by for some earache for daring to mention that.

Go speak to any HVAC engineer. They will all regale you of tales of mental torture in trying to get a ambient room temp satisfactory to all. It's an impossible task.

Been there I used to joke that in an office with 10 women, you'd get 11 opinions as to desired temperature. Not only that, but demands that there be no air movements - nor when cooling, any cold air.
I'll also add the tendency fir the stick insects to arrive in the depths is winter, in light summer clothes, get chilled to the bone getting from the car to the office, and then complain they are cold. And if you dare point out that they could wear something more suited to the weather ...

My neighbour uses his thermostat as an on-off switch. explaining to him how it works is an annual ritual.

Had that as well. It appeared to be an innate inability in many people to grasp the concept of a thermostat regulating temperature.
Where I worked, we had to remove all the user accessible controls for this reason.

How to cope with all that input into a control system for a slow reacting heating system?

At the risk of a slight loss of efficiency, a buffer helps. In my case, I went with a thermal store for a number of reasons, but the decoupling of the heating system (inherently variable flow rate) from the requirements of the boiler (hard minimum flow rate & power output) was one of them.
It also decouples control - the boiler is run just from a stat in the cylinder and is not connected to the CH controls. Though it's probably not as simple with a heat pump which ideally needs to operate only at the temperature needed for the heating.

 

[MNW67]

But on zoned systems ( especially with underfloor) when a zone is being heated and then another asks for heat the return temp drops quickly and in most cases the heat pump either defaults to minimum putput or goes into a waiting period.

You mentioned this happens because of the heat pump's internal logic, which is something I am still learning about. Why does a big drop in return temperature make the heat pump go to minimum or waiting period? I am trying to work out what it assumes is happening and what it is trying to achieve.

 

[MNW67]

If you have massive thermal masses and good insulation, then there will be limited drop in temperature, and so little saved in heat losses. At tge other extreme, the temperature reduction and hence heat loss would be significant.

I am still trying to get my head round all of this and have been looking into Decrement Delay.

This takes you into a detour of looking at the "Decrement Delay" of the building:

Isn't the effect of large thermal mass to divorce heat losses from temperature changes? So, you still have the heat loss but the temperature doesn't actually change very much. But conversely, when you reheat, you need to put a lot of heat in without raising the temperature very much.

 

[Ivor Windybottom]

I reckon you can over-analyse. Despite my engineering design background I didn't bugger about with maths at all, I used HeatPunk (free!), you just draw your house, tell it what each wall is composed of then it helps you to find the right radiator sizes. You can then play games with flow temperature vs radiator size.

I've ended up with a design that's tolerable from an aesthetic angle that will theoretically heat the place adequately at a flow temperature of 42 degC.

I'm going to keep quiet about all this homework when I start talking to plumbers soon. I don't want to be the nightmare smartarse customer, but I will quietly compare their design with mine. Unless they come up with something better of course.

I'm just going to install the thing and see what happens. The zone valves will be installed but all manually locked open. Then I can start manually closing them and see what happens. I'm also bench-testing timer/thermostats for each of the zones in parallel. Hopefully I'll arrive at a point where I have a good plumbing system and controls I'm happy will do the job, then I just need to wire them in.

I promise to post my findings!