Posted by: @heatgeek

↑

As an example for a low energy house with UFH and grossly over-sized heat pump, a 300l independently heated (by HP) low temperature thermal store with 8C minimum DT low flow rate to thermostatically controlled rooms will achieve much better energy saving than any form of open-loop working,

Are you sure about that. I have done almost every configuration of heating possible including what you describe,it isn't the cheapest or most comfortable.

By far the lowest cost is batch charging direct to floor in cheap period, if without battery, so around 7 hours overnight. That is open loop direct to floor (you a single low hysterisis thermostat for that). Heat pump should run continuously for whole period and ASHP generally be well oversized.

If you have a battery run elevated flow temperature for 7 hours then normal WC (in a low energy house the ASHP should stay off for nearly the whole day, when really cold may start again early evening - again all open loop. 

Neither of the above get any advantage from the thermal store. Especially if using a dT of 8.

If you have battery and PV, just run heat pump at elevated temperature on excess PV once battery is nearly charged, then it's free heating. Most the shoulder seasons can be done this way. Then back to straight WC. When we did this in Feb and March, night temps were dropping to zero and the floor remained hot enough to keep heat pump off.

Running 8 dT, means heating your thermal store hotter than needed, compared to dT of 4, for the same mean flow temperature in to floor, so isn't the correct way to go. And thermostat in all rooms, been there also, not the cheapest for low energy house 

Are you sure you have a low energy house?

Posted by: @heatgeek

↑

Large solar gain

Are you sure you get large solar gains, we can get temperature close to 30 in the living room, if we let the solar gain run uncontrolled.

SCoP is pretty meaningless cost is all important. I am currently making my scope worse each day as I am now in cooling mode getting a negative CoP.

Posted by: @damon

↑

id use a buffer when the job would dictate.for example on a 13/16 Viessmann vitocal you don’t get much head left on the pump, so a buffer would allow separation and additional pump.

id usually go with a 3 pipe buffer on the flow, upsize the tee and entry point init the buffer.

...

you also need inside your buffer correctly, generally you find issues with smaller buffers and header where low volume affects the temps more severely and also the set up, the little buffers under cylinders are useless.

Thanks for that thoughtful reply. 

Unfortunately we get far too many reports here where 50l 4 port buffers have been installed 'because that's what we do' often, so far as we collectively can tell (remotely), contributing significantly to poor performance which is why many on here will advise against them unless there is an identifiable reason.

It sounds, in contrast, like you 

• Only use a buffer when there is an identifiable reason
• If you do use a buffer, take several  careful steps to reduce system distortion

Is this a correct reading of your approach.

Posted by: @damon

↑

you could also use a buffer where you have different circuits and the customer wants control over rooms.

Is there any reason in not to use trvs (or zone valves) without a buffer in the rooms which need (and can sustain) control (usually a minority).  That's how my system is set up and it seems to work well. 

As ever I'm thinking about ordinary houses (typically 100sq. m in the UK, but my comment is intended to apply to houses up to perhaps 250-300sq m which are not very long and thin).   I can see that with a very extensive and complex system multiple pumps and a buffer might be useful.

Posted by: @damon

↑

Posted by: @damon

↑

scop is all well and good but we should also be looking at energy usage, you could be getting a better scop but using more energy.

 

 

Well said, Sir! I have been trying to get this point across to reluctant ears.

I'm curious who you think are the reluctant ears.  Certainly not me, and if you do think i am one of the people you refer to, then you are misunderstanding my comments! 

Nor have I seen many others suggesting what you say, with the very important exception of those who go to the next step and look for lowest cost using tou tarrifs, often as a result consuming more energy than would otherwise be the case.

↑

Not a buffer or volumiser in the normal sense. The buffer only gets used during defrosting, heating the buffer up prior to defrosting to a high temperature, then using only the energy within the buffer during defrost. It is then isolated. Seen results of this arrangement and the CoP overall was better than the normal arrangement.

The water pump is the circulation pump for ASHP, just located inside not outside.

We pushed for a system without a buffer as per advice of many in this forum. The Arotherm fits this and with a heatloss of 5.1, the 5KW version was tempting. But I am hearing that some leeway is needed, like this nearby heatpump user shared: 

One word of advice. All heat pumps pause to run defrost cycles at outdoor temperatures below about 3C which reduces their overall output more than the -2C specification might indicate, so make sure that your heat pump is rated to more than your house requires. For instance my house demand is 5.5kW but I find that my 7kW heat pump just covers it at sub-zero temperatures.

I had the sense that, in the SE, defrost would not run often. But if defrost starts around 3C, this may be more of a factor?

@batpred 

Many manufacturers sell only one or two models of heat pump, then add software limits to make intermediate power models. The result is, if you buy a "5kW" heat pump which is actually a 7kW machine limited to 5kW, then the lower end of the modulation range will be exactly the same. Only the maximum power is different. When we say "don't oversize the heat pump" we're mostly worried about it cycling a lot because the lowest possible power at the low end of the modulation range is too high. But if the two machines you're hesitating between are actually the same, then... why not pick the bigger one? That will extend the modulation range upwards, but the minimum power will be the same. That may still cause problems if there's not enough flow, though.

On the other hand, if the manufacturer makes 3-4-5kW machines from the same chassis, and 7-9-12kW from another chassis, then the 5 and 12kW models have full modulation range allowed by the hardware, but the limited models have the low end of the modulation range from the bigger heat pump!

I'm not sure how Vaillant splits their models, you should check.

Posted by: @batpred

↑

One word of advice. All heat pumps pause to run defrost cycles at outdoor temperatures below about 3C which reduces their overall output more than the -2C specification might indicate, so make sure that your heat pump is rated to more than your house requires. For instance my house demand is 5.5kW but I find that my 7kW heat pump just covers it at sub-zero temperatures.

I had the sense that, in the SE, defrost would not run often. But if defrost starts around 3C, this may be more of a factor?

That word of advice isn't entirely accurate.

The heat pump will run a defrost cycle whenever it starts icing up. That is certainly likely at outside temperatures approaching or going below zero but it's also perfectly possible at higher outside temperatures too. More relevant is how moist the outside air is, since there'll be relatively little icing up at (for instance) -5 if there's almost no humidity but probably quite a lot of icing up on a misty day at higher temperatures with lots of water in the air.

The South East is, as you know, comparatively warm and comparatively moist, so heat pumps will definitely go through their defrost cycles quite regularly in the cold months. I have certainly seen my heat pump go through defrost cycles at OAT greater than 3degC and seen plenty of times where there's been almost no defrosting at subzero temperatures. In other words, it's nuanced.

Posted by: @bobflux

↑

When we say "don't oversize the heat pump" we're mostly worried about it cycling a lot because the lowest possible power at the low end of the modulation range is too high. But if the two machines you're hesitating between are actually the same, then... why not pick the bigger one? That will extend the modulation range upwards, but the minimum power will be the same. That may still cause problems if there's not enough flow, though.

Thanks, I had a look at https://energy-stats.uk/heat-pump-cycling/ and the lower modulating is shown as 2.2 for the 5kW arotherm vs 3.2 for the 7kW. 
If the 5.1 kw loss is confirmed and a 5kw pump recommended, I would need to check whether the advice below would apply to it.. 

For instance my house demand is 5.5kW but I find that my 7kW heat pump just covers it at sub-zero temperatures.

I am an Alberta, Canada installer new to A/W heat pumps.  I just signed up to distribute Arctic Heat Pumps, which I believe sell rebranded Macon R32 inverter units.  My supplier says I have to install a buffer tank .. yet the design I want to use is a simple, 2 zone .. full flow design with a massive concrete hydronic slab for one of the zones, an air (ducted water to air exchanger) for the second loop, and no DWH loop.  My supplier says what works in Europe won’t work here because we are much colder.  This doesn’t make sense to me.  The HP from Arctic appears to be designed to be solely controlled by a buffer tank temperature sensor plus ambient sensor.  I want to use the concrete slab as my energy mass for defrost.  Maybe the problem is the HP they sell is too crude to be used without a buffer?  Any advise?  I have 2 question here … One .. can buffer-less work in cold regions like Alberta? .. and two .. Do I need to find a different HP supplier? (Ie a HP with more advanced / different controls) … I can’t see how to wire this Arctic unit without a buffer?  Thanks! 

Posted by: @tony-stolz

↑

I am an Alberta, Canada installer new to A/W heat pumps.  I just signed up to distribute Arctic Heat Pumps, which I believe sell rebranded Macon R32 inverter units.  My supplier says I have to install a buffer tank .. yet the design I want to use is a simple, 2 zone .. full flow design with a massive concrete hydronic slab for one of the zones, an air (ducted water to air exchanger) for the second loop, and no DWH loop.  My supplier says what works in Europe won’t work here because we are much colder.  This doesn’t make sense to me.  The HP from Arctic appears to be designed to be solely controlled by a buffer tank temperature sensor plus ambient sensor.  I want to use the concrete slab as my energy mass for defrost.  Maybe the problem is the HP they sell is too crude to be used without a buffer?  Any advise?  I have 2 question here … One .. can buffer-less work in cold regions like Alberta? .. and two .. Do I need to find a different HP supplier? (Ie a HP with more advanced / different controls) … I can’t see how to wire this Arctic unit without a buffer?  Thanks!

Firstly well done for questioning the statement...What's the purpose of the buffer?  Also what exactly is meant by a buffer.  Those are the questions to ask.

Here in the UK a buffer is generally taken to mean a 2, 3 or 4 port tank which connects flow to return.  They can work but all too frequently, with UK installation practices, compromise performance.  A volumiser on the other hand is generally taken to mean a 2 port tank in either flow or return.  These do not compromise performance.

If the issue that the 'buffer' is trying to solve is defrost, as you hint at but don't confirm, then a volumiser will work just as well without the downside (as will a large slab).

If the issue that a 'buffer' is trying to solve is an unavoidable need for different flow rates in primary and secondary, then a buffer, phe or llh may indeed be necessary.  But before you conclude it is you should examine why different flow rates are needed and what are the knock on impacts.

As with any engineering question you need to understand the problem before you can design a solution.

Hope that helps, feel free to ask for further clarification.  I recommend you now go back to my first paragraph ask the questions posed. If your supplier or the OEM can't answer them then I suggest you need to find someone who understands the product they are selling!  You don't want to be in the position of reselling a product which your supplier doesn't understand.

One final question, why R32.  It's being phased out in the EU because of high gwp and R290 performs better in many respects,  and has low gwp.  R290 pumps gan reach fts of 75C which means they can do a legionella cycle natively and the higher temp can occasionally be useful for extreme circumstances.

@jamespa Thanks for the quick reply.  My supplier is suggesting a 4 port buffer for 3 reasons .. to assist defrost, to reduce short cycling, and to ensure there is no flow restriction.  I have already done the math .. including heat loss, heat / cool demand, sizing, plus all flow rates for the system.  I know my 2 zones will accept full flow.  I know I have tons of mass in the slab for the defrost cycle.  I do math for a living .. which is why I am questioning this … my math/physics brain says I don’t need a buffer!  

Bottom line .. you answered my first question.. which I thought was obvious as well .. I don’t need a buffer.  I wouldn’t mind your opinion on the second question though … the HP that my supplier offers has very simple controls in that it only uses a buffer tank sensor plus an ambient temperature sensor .. plus whatever DT you enter in the control unit to modulate output.   How can this design even work without the buffer it was designed for?  I’m thinking the pushback I’m getting from my supplier is because the HP they sell can’t be used without a buffer.

Re R32 vs R290 .. yes I’d love to bring in an R290 unit as they have higher COP, work better in cold conditions , and have lower GWP .. my government unfortunately doesn’t allow it because they are Luddites (they are concerned about the teapot size of flammable propane in it).  We really don’t have great options here … our grid is 60 hz .. so it isn’t like I can import a 50 hz unit from Europe.  

I am very happy I found this forum … it is next to impossible to have these conversations with anyone here.  BTW .. I am an expert in ROI .. thus my focus on simplicity and reducing the capex of the systems I design.  I just need to find a HP brand that I can get here and some instructions / wiring diagram on how to install it the way I want.  

Posted by: @tony-stolz

↑

the HP that my supplier offers has very simple controls in that it only uses a buffer tank sensor plus an ambient temperature sensor .. plus whatever DT you enter in the control unit to modulate output.   How can this design even work without the buffer it was designed for?  I’m thinking the pushback I’m getting from my supplier is because the HP they sell can’t be used without a buffer.

Hopefully the ambient temperature sensor drives a weather compensation curve, ie a function that varies the flow temperature with outside temperature.  This is always the starting point for operating a heat pump efficiently and with greatest comfort and for many people its all you need, because by adjusting flow temperature you adjust the amount of heat emitterd and if you get the curve right then that just balances the house loss (which of course is a function of outside temperature

Most heat pumps will also have a room temperature sensor which can be used in various ways.  What you dont normally do is use a thermostat in on/off mode to control, particularly with a large slab where doing that will almost certainly result in oscillatory behaviour.  

No idea what the buffer sensor does, its probably on/off in which case could probably be short circuited so its always on, or if its a thermistor replaced by a fixed resistor so its always showing too low.  Possibly it can be disables in the controller.

If the heat pump dosn't support weather compensation then dont touch it.

I suggest you read this simple guide to heat pumps which might put things in a bit more context.  Forget almost everything you think you know about running boilers (do you call them furnaces in Canada or is that a US thing), much of it was wrong for boilers and it is definitely wrong for heat pumps.

@tony-stolz, I think @jamespa has already covered the hydraulic fundamentals extremely well, particularly the distinction between genuine hydraulic separation and simply adding water volume, so I won’t retread too much of that ground.

My own view is that the “you must have a 4-port buffer because you’re in Alberta” argument is too simplistic. Cold climate alone does not suddenly invalidate bufferless design principles.

That said, I do think extreme climates strengthen the argument for stable system volume and hydraulics generally. In Alberta, you’re dealing with longer periods at very high compressor load, harsher defrost conditions and much less margin for unstable flow behaviour than we typically see in the UK. So there is a legitimate engineering discussion around thermal mass, guaranteed flow and anti-cycling in those environments.

But that still doesn’t automatically justify hydraulic separation via a 4-port buffer.

Your slab already is a massive thermal battery. It provides real usable thermal mass directly connected to the emitters rather than an artificial tank inserted into the system to compensate for uncertainty elsewhere in the circuit. IMO that’s thermodynamically cleaner than a traditional buffer arrangement.

Where I think the real issue lies is product philosophy (if that’s the right word to use).

A lot of modern heat pumps are engineered around installer protection as much as system optimisation. The buffer becomes an “anti-problem device”. It protects the appliance from poor hydraulics, zoning instability, uncertain water volume and inconsistent commissioning. From a manufacturer’s perspective, that makes complete sense because it reduces callbacks and protects reliability across a broad installer base.

The downside, as we’ve repeatedly seen in the UK, is that once you start designing systems around protecting the heat pump from the installer, compromises creep in everywhere. Buffers, LLHs and secondary pumps often end up masking design weaknesses rather than solving them.

And honestly, your proposed setup sounds more coherent than many modern systems I see over here discussed by installers. Large slab, continuous circulation, known flow rates, stable low-temperature operation and huge thermal mass available for defrost energy… that’s exactly the kind of environment heat pumps tend to like.

Where I do think your concern is valid is the controls side. If the Arctic/Macon unit fundamentally expects to “see” and maintain a buffer tank temperature, then you may end up fighting the internal logic of the machine. At that point, the issue isn’t Alberta and it isn’t defrost… it’s whether the product architecture supports direct-to-system low-temperature hydronics. I also wonder how they monitor that.

Your ROI mindset is probably the right one. Simplicity nearly always wins long term. Every additional separator, buffer, pump, plate exchanger or hydraulic workaround adds parasitic losses, commissioning variables, control complexity and another future failure point.

So my own stance would be this… cold climate strengthens the need for stable volume and hydraulics, but I don’t think Alberta itself mandates a 4-port buffer if the slab already provides substantial thermal mass and the system flow is properly engineered.

What may actually be forcing the issue is the specific heat pump’s control philosophy and the manufacturer trying to create a universally installer-safe installation envelope. That’s a very different conversation from saying the building physics require it.