Once again, thank you very much folks.

Overall, I’m hearing that there are ways to mitigate the different disadvantages of the two options, but probably less so with the Grant.

In very cold weather, even if the Grant 10kW is kept running 24 hours a day it will still struggle to reach the desired temperature, due to the issues mentioned by @allyfish with defrosting.

With the Samsung, on the basis that that it’s oversized, we’re likely to find ourselves in a similar situation to @peterrsaltings, and could expect similar results to him with his approach of timing the HP to run on and off; to be clear, this effectively eliminates the problems with the Samsung not modulating down below 4kW, and short cycling on and off; it also means the house isn’t over-heated.

@iancalderbank, you say “if you adjust it to lower than what your radiators have been sized for, you'll be cold. but it’s also possible that if the heat loss has been over-estimated, then so has the design flow temperature.” However, with the Samsung, if I reduced the flow temperature from the design temperature of 50c, would this just exacerbate the problems of oversizing, because it’s even more underemployed, with even less power required. I’m wondering how adjusting for a lower flow temperature would interact with implementing Peter’s ‘long cycling’ approach – would it get in the way of running that successfully?

And another thing to check @iancalderbank: you say “Samsung has a swing of 1C (you can't change it)” –  I’m not fully understanding whether this could also get in the way of implementing the long cycling approach, of controlling the timing of when it is on and off and through that achieving long cycles?

@tobyg the 1C swing between on and off is native to the samsung controller when used as thermostat. if you want a different window to that, you need to use a 3rd party stat with the right characteristics..

strongly suggest you get your radiators resized for 40C not 50C. its a fundamental thing that almost everyone whose heat pump doesn't work as well as they'd hoped says: I needed bigger radiators. when you lower the flow temp it lowers the radiator power. so you need bigger radiators to maintain the same overall power. adjusting the flow temp back up again is a workaround that gets more power out of the radiators you have  but at the expense of inefficient use of electricity and significantly worse performance (and possibly actually not a warm enough house) in sub-zero defrost conditions.  This is irrespective of which brand you choose.

@iancalderbank, on this point about the 1C swing between on and off with the Samsung I was wanting to check whether this has any bearing on operating the long cycling approach set out by @peterrsaltings – part of what he’s saying is that the long on and off cycles involve allowing the room temperature to swing back and forth by around +/-2c, so I was a bit confused in case you were referring to something that would limit that. I think you’re talking about something different, or are we saying that to implement @peterrsaltings approach requires a different type of thermostat / controller? I’d like to be clear on this, because if we get the Samsung we’d probably be relying on this approach – could you confirm please?

I’m trying to now get my head round flow temperature and radiator sizing, but what I’m hearing is that this is a different issue, and the extent of any problems here isn’t affected by whether the HP may be over or undersized (i.e. Samsung 12kW or Grant 10kW). But what I do want to check now is in the event that we go for the Samsung, but the installers won’t resize the radiators for a lower flow temperature, would that have an impact on our ability to implement long cycling?

We do know the proposed radiator sizings, but not the type – I will ask. When I plug these sizes into the Freedom HP spreadsheet, and assuming for now that they are double convectors, most of them come out as big enough by some margin, however, the spreadsheet doesn’t take account of flow temperature.

Posted by: @peterrsaltings

↑

As I said, the drawback is that on days like today, when the ambient is around 10-12C, day and night, the heat pump was only on for 3 or 4 hours and most of that time was modulating down to 1KW input (the lowest it can go), with 4-5KW being generated. The compressor is a bit noisy when doing so.

Thanks so much for your long and considered post yesterday Peter - for some reason I've only just seen it, and am working my way through what you say, and noting that most of it is pretty positive. Just on the noise though, this is a bit of an issue for us, if part of what happens when it's down at it's lowest is that it's noticeably noisier.

ok, @allyfish, @iancalderbank,and @editor, as you’ve advocated, I've finally managed to produce a heat loss calculation, using the Freedom Heat Pump spreadsheet recommended by @hughf, and I admit it, it wasn't as difficult as it first looked (though it seemed thorough). This comes out at 9.7kW, so very close to the installers' figure. And here's the range from the various methods I've tried:

(It’s worth mentioning that one large room in particular pushes the heat loss up considerably, accounting for 3kW (in the installers calculation) - this is a single storey extension, insulated, but with 3 external walls, a pitched roof, and lots of windows; we currently use a oil-filled radiator in there and would be happy to continue with that as a supplementary heat source.)

This Freedom heat pump calculation begs some questions – have the installers got it right after all, including with their initial proposal for a Samsung 12kW, and have we been worrying needlessly about it being oversized? Ought this to steer us away from the less powerful Grant? Or does the heat loss figure based on current gas usage, way down at 3.2kW, still tell us something important here?

I presume part of it is that the Freedom HP calculation takes a similar approach to MCS – that the house needs to be up near 21c, in every room, all the time, which is not how we’d probably run it. But still, it's a big discrepancy in the figures above.

Now we’ve got these figures, if anyone here is willing to give their final opinion on which you think we should go for – Samsung 12kW of Grant 10kW – please don’t hold back!

Thank you.

My personal opinion: Samsung. Grant unit is a good workhorse, and my calculations of the capacity it supplies are approximate, but I’ve never observed it kick out more than 9kW flat out.

@tobyg well done on completing the heat loss calculations.

When it comes to selecting the right heat pump, it’s not just about hitting the exact number but understanding the conditions and how you use your heating.

The close match between the installer’s figure and the Freedom Heat Pump calculation suggests that the installers' recommendation may be well-founded, especially considering the specific high heat loss from your large room. However, the lower figures based on actual gas usage can't be ignored as they reflect real-world usage and may indicate that your overall heating demand is less than theoretical calculations suggest.

Given your supplementary heating for the high-loss room, you might lean towards the lower capacity option, but consider this: Will the Grant 10kW suffice on the coldest days without the supplementary heating, and is that supplementary heating more cost effective than running everything through the heat pump? If yes, it could be a more efficient choice in the long run.

The discrepancy does suggest a conservative approach by the installers, likely to ensure comfort during the coldest periods. Your usage patterns and comfort levels are essential to consider – if you don't need every room at 21°C constantly, the lower figure might be more reflective of your needs.

Having said that, based on @allyfish’s assessment above pertaining to the Grant, I’m with him on the Samsung. 

Posted by: @tobyg

↑

@iancalderbank, on this point about the 1C swing between on and off with the Samsung I was wanting to check whether this has any bearing on operating the long cycling approach set out by @peterrsaltings – part of what he’s saying is that the long on and off cycles involve allowing the room temperature to swing back and forth by around +/-2c, so I was a bit confused in case you were referring to something that would limit that. I think you’re talking about something different, or are we saying that to implement @peterrsaltings approach requires a different type of thermostat / controller? I’d like to be clear on this, because if we get the Samsung we’d probably be relying on this approach – could you confirm please?

I’m trying to now get my head round flow temperature and radiator sizing, but what I’m hearing is that this is a different issue, and the extent of any problems here isn’t affected by whether the HP may be over or undersized (i.e. Samsung 12kW or Grant 10kW). But what I do want to check now is in the event that we go for the Samsung, but the installers won’t resize the radiators for a lower flow temperature, would that have an impact on our ability to implement long cycling?

We do know the proposed radiator sizings, but not the type – I will ask. When I plug these sizes into the Freedom HP spreadsheet, and assuming for now that they are double convectors, most of them come out as big enough by some margin, however, the spreadsheet doesn’t take account of flow temperature.

we're talking about the same thing. If you want to allow your room temps to vary by +-2c you'll need a 3rd party thermostat (thats a 4C swing). Not sure anyone would want that wide to be honest. The samsung controller used as thermostat is +1C -0C , i.e. set it to 21, it'll keep the house between 21 and 22.

yes if your radiators are undersized and your HP is oversized you are more at risk of getting short cycling in mild conditions. thats why I keep saying push for biggest rads you can. you're perfectly entitled to them giving you full radiator spec. post them here. if possible get their power ratings at DT50 (the "boiler standard" rating) as well, converting that down to lower flow temps is just a formula.

Posted by: @tobyg

↑

I presume part of it is that the Freedom HP calculation takes a similar approach to MCS – that the house needs to be up near 21c, in every room, all the time, which is not how we’d probably run it. But still, it's a big discrepancy in the figures above.

The Freedom HP calculator allows you to put different desired room temperatures in different rooms. If you change from 21C to what you would run your rooms at, what heat loss figure does that come to? My heat loss calculation using 18C in most rooms is very similar to yours - approx 10kW. If my heat meter is to be believed, my actual heat loss is closer to 7kW. This is based on measuring OAT, IAT and heat production from the heat meter when it is very cold outside and the heating is balancing the apparent heat loss. My short cycling problems with my 12kW Samsung seem to have resolved with using room temperature mode, together with a higher LWT and allowing the wired controller thermostat to switch the ASHP on and off. The figure shows today's ASHP activity based on wired controller thermostat set on 16C from 00.30 to 05.00 and 17C from 05.00 to 07.30, then 16C from 07.30. The wired controller reads room temperature around 0.5C lower than my other temperature sensors in the same area, so we are not quite as Spartan as it seems. I haven't noticed it cycling even when OAT was 11C and LWT was 33C.

Posted by: @tobyg

↑

We do know the proposed radiator sizings, but not the type – I will ask. When I plug these sizes into the Freedom HP spreadsheet, and assuming for now that they are double convectors, most of them come out as big enough by some margin, however, the spreadsheet doesn’t take account of flow temperature.

Bear in mind the recent post about Radiatorgate ie quoted radiator outputs may be overoptimistic. Also, any buffer or low loss header may, depending on its design, reduce the flow temperature to your radiators by 3-4 degC. The last factor may be peculiar to my retrofit set up, but my furthest radiator is 4C cooler than the nearest radiator.

@tobyg

I have attached a copy of a Radiator Heat Output predictor which you may find of use for both individual radiators and the heat emitter capacity for the whole home.

The yellow highlighted cells are values that can be changed, with the predicted thermal energy output being shown in the green highlighted cell located in the left table. The right table can be used to predict the probable required average water temperature.

As an example with a LWT of 45C and a RWT of 40C, the average water temperature at the heat emitters will be 42.5C. If the total capacity of all the heat emitters at Delta 50 is 30000W, then the predicted thermal energy supply to the home will be 10015W at an IAT of 21C.

I hope that you find this of use.

With regard to heat loss calculations, spreadsheet based calculators deal only with building fabric, which of course does not take account of other factors. Human activity within the home will be one source of additional thermal energy, possibly in the order of 10kWh per day. Probably the largest unmeasured source of additional thermal energy is solar gain, not just though the windows but also on the external fabric of the building. By solar gain I don't mean changes in OAT, but actual heating of the external fabric of the building, which of course will reduce the temperature differential between inner fabric and outer fabric, thereby reducing the actual heat loss. Wind chill and rain effect may of course have the opposite effect.

@tobyg Just to buck the trend, I'd go for the Grant. Now, I have a Grant 10kW and it's proved quite tricky to set-up and get what's needed out of it, but I'm only one winter in, and have learned quite a lot.

Here are my points:

1. The Grant functions well in mild temperatures; it'll happily produce output under 3kW (in fact, under 2kW) for hours on end without having to cycle on and off. As an example, a mild February day, running for 8 hours without turning off, requiring 1kWh/hr:

(power is in multiples of 100W)

(obviously results will vary from house to house)

2. On cold days, below 3C, I've needed some supplementary heating (a 700W electric heater) to get the house up to heat, but once there, the Grant will maintain that temperature, even with defrosting. So far there have been 26 days where the average OAT was below 3C. Consider that I live in a village in Cumbria = cold and wet; you might be in a milder part of town. 

3. When running slowly (i.e. modulating down to its lowest output levels) it's very quiet. The harder it works, the more sound the outside unit makes, but when it's in its happy "slow and low" state it's inaudible from inside, and the unit it directly outside my lounge window, under 2m from where I'm currently sat. I sometimes open the window to check it's still on. Maybe I'm deaf.

4. From a simple "save the planet" point of view, going for a lower powered model will be a greater incentive to ensure insulation is covered well - make sure the loft is filled, look at leaky doors and windows, and so on. The more you address this, the lower your energy requirement (and demand, and bills) will be. I sometimes wonder, if I'd done all these things when I had a gas boiler, how much could I have saved? But then. if you have heat energy to burn (literally) it's easier to overlook.

Looking at the forum roll call, there's around 20 Grant owners, and about 12 Samsung owners. Small sample, and I guess you could read that two ways: the Grant is more popular, or the Grant is more troublesome (as people tend not to seek out forums when things are going well).

The Grant's current controller is poor, and hopefully you'd get at least the new Grant smart controller. If not, ask for one. Whatever the Samsung has, can't be as bad.

And final point on the Grant - you'll probably have read many things about the Grant LLH, and it's one of the main sticking points in recommending it. I'd probably have preferred an installation without it, but that's a hard sell, as an installer won't want to go against the manufacturer's recommended installation path (although Grant might be shifting on that stance from some items I've read). However, if you have to have the LLH, try to get it wired up on installation, as it might (although there's no definitive answer it seems) help reduce defrost cycle pain. Certainly it'd be more convenient to have it wired up at installation than after. It's only 1 extra cable from the unit, and power to the LLH.