Buffer control

[SuperJohnG]

I'd followed a topic for a specific reason as someone gave a good hint on controlling their UFH buffer - I think it was just that in addition to the call for heat from the buffer (turning on the ASHP) they also wired the call for heat from the UFH that way the buffer wasn't getting heated by the ASHP all the time if you weren't using the heating.

 

Despite going through18 pages of followed content and using every search term I can think of ..I can't find the post. 

 

So two questions: 

 

Does anyone know of the post? 

 

Secondly how do you control your UFH  buffer? 

[JohnMo]

Two ways to control

1 attach a thermostat to buffer, run it hotter than you run ch use the thermostat to call for heat from ASHP.  Disadvantages, lower CoP from heat pump, cannot run WC.

2 allow buffer to float on ASHP flow temp. Can use WC keep flow temp low. Use room thermostat to call for heat.

 

If you haven't plumbed it in yet, consider piping into the return line only, not across the flow and return this ensures no mixing of supply and return flows to minimise flow temp. It's also heated to the lowest temperature so least standing losses.

[SuperJohnG]

It's not piped yet. I didn't know you couldn't use weather compensation when heating the buffer? 

[JohnMo]

You can't if the buffer is controlled by a thermostat.

 

If your running WC can you do away with the buffer by using a 1 or 2 zone system, just make sure the engaged volume is bigger than minimum volume required by HP.

[Nickfromwales]

6 hours ago, SuperJohnG said:

I'd followed a topic for a specific reason as someone gave a good hint on controlling their UFH buffer - I think it was just that in addition to the call for heat from the buffer (turning on the ASHP) they also wired the call for heat from the UFH that way the buffer wasn't getting heated by the ASHP all the time if you weren't using the heating.

 

Despite going through18 pages of followed content and using every search term I can think of ..I can't find the post. 

 

So two questions: 

 

Does anyone know of the post? 

 

Secondly how do you control your UFH  buffer? 

You wouldn’t have the ASHP firing without demand downstream of the buffer?! 
Call for heat would fire both the primary ( ASHP ) and secondary ( pump at buffer > manifold if required ) simultaneously, and when the buffer gets heated to the set point of its stat the ASHP will rest, but the UFH will still continue to pull heat from that body of water ( buffer volume ) until the buffer stat kicks in again, and repeat.

Call for heat goes away, everything goes to sleep. Simples. 

[SuperJohnG]

2 minutes ago, Nickfromwales said:

You wouldn’t have the ASHP firing without demand downstream of the buffer?! 
Call for heat would fire both the primary ( ASHP ) and secondary ( pump at buffer > manifold if required ) simultaneously, and when the buffer gets heated to the set point of its stat the ASHP will rest, but the UFH will still continue to pull heat from that body of water ( buffer volume ) until the buffer stat kicks in again, and repeat.

Call for heat goes away, everything goes to sleep. Simples. 

 @Nickfromwales it was only so long before I tagged you. Thanks for coming back.

 

I am unsure if you mean that's correct? 

 

Room stat calls for heat - UFH to buffer pumps starts, and ASHP starts IF buffer stat is lower than setting (say the buffer stat is set to 35 degrees).  

If the buffer was already at temp required, I.e. buffer stat happy temp has been reached then the ASHP wouldn't need to start? 

 

But say in the summer - if you don't want the buffer constantly heated, (when not using the UFH) to its stat temp then it doesn't turn on if there is no call for heat from the room stat. Downside being it would take a while to heat the buffer (no idea how long yet) as its off when no call for heat from the room stat. 

 

BTW- just took delivery of my IVAR blending set which I'll set the temp down on to say 25 degrees. 

 

I have pipes at 200mm centres, single zone on a 265m near passivhouse.

 

[JohnMo]

If your on a single zone at 200mm centres and 265m2, don't see why you need a buffer. Unless your running a massively oversized heat pump.

 

Mine (6kW) will be installed with a single zone, but with 300mm centres on 192m2, with a second area of 16m2 on 115mm centres (summer house) - no need for a buffer.  All will be run on WC.

[SuperJohnG]

50 minutes ago, JohnMo said:

If your on a single zone at 200mm centres and 265m2, don't see why you need a buffer. Unless your running a massively oversized heat pump

I wanted a buffer to avoid short cycling. 

 

I haven't chose the heat pump yet. I'm currently trying to do that, but struggling to get a good answer out any suppliers. 

[JohnMo]

2 hours ago, SuperJohnG said:

wanted a buffer to avoid short cycling

 

Not sure why you think you will get short cycling when operating as a single zone.

 

A HP that turns down to say 3kW and has a 0.5kW heat demand, requires a system volume of 60L to give a run time of 10 minutes.  Off time should be around 50 minutes.

[SuperJohnG]

40 minutes ago, JohnMo said:

 

Not sure why you think you will get short cycling when operating as a single zone.

 

A HP that turns down to say 3kW and has a 0.5kW heat demand, requires a system volume of 60L to give a run time of 10 minutes.  Off time should be around 50 minutes.

Just based on reading here, albeit I am currently doing more reading and it would seem maybe I don't need a buffer. 

 

I'll have about a 70l volume. 

[Conor]

@SuperJohnG you don't need a blending valve. You set the better set temp to a couple degrees above what you need for the UFH. Mine is set to 34c to deliver 30-32c at the manifold.

 

My setup is roughly:

 

Flow from ASHP goes to plant room, three way valve then splits flow to either the HWC or buffer.

Return pipe from HWC and buffer connect and return to the ashp. 

Pump on the buffer sends flow to the UFH manifolds via 28/22mm pipe. 

 

Controls.

Hot water demand is set by the DHW stat and the ASHP controller. Nothing to do with the UFH.

Heating demand comes from the wiring centre, sends signal to ashp to start running, opens valves depending on zone calling and starts the main pump and appropriate buffer pump.

Ashp heats water and modulates to the stat temp in the buffer, ramping down as target temperature approaches.

 

As ahsp don't "fire" when your UFH stats call for heat, the ashp wake up from standby regardless and start getting pumps going etc etc. At least mine does. So it's always pumping to the buffer when there is demand. Don't know if all ashps do this tho .

 

You'll not need the heating past March or so, heating up time of the buffer isn't an issue.

 

[dpmiller]

^ to avoid any ambiguity, @Conor has a sensor in the buffer that was supplied with the ASHP, some brands don't have this option.

 

I'd suggest if you want to have a thermostat on the buffer, simply putting it in series with the demand signal from the wiring centre might work for your case? If either heat demand or buffer are satisfied the the ASHP will be commanded off, and it'll only come back on again if you still need more heat *and* the buffer has cooled.

[Nickfromwales]

13 hours ago, SuperJohnG said:

Room stat calls for heat - UFH to buffer pumps starts, and ASHP starts IF buffer stat is lower than setting (say the buffer stat is set to 35 degrees).  

If the buffer was already at temp required, I.e. buffer stat happy temp has been reached then the ASHP wouldn't need to start? 

Correct. 
 

13 hours ago, SuperJohnG said:

But say in the summer - if you don't want the buffer constantly heated, (when not using the UFH) to its stat temp then it doesn't turn on if there is no call for heat from the room stat. Downside being it would take a while to heat the buffer (no idea how long yet) as its off when no call for heat from the room stat. 

In the summer your heating will all be in the summer / off position, so I don’t get why your system would come on, to heat anything? 
Summer sees the ASHP > UVC directly, and the buffer remains isolated and dedicated to the heating primary circuit only. 

[SuperJohnG]

2 hours ago, Conor said:

@SuperJohnG you don't need a blending valve.

The one at the manifold, it's got the UFH pump on it and a blending set so the floor can never go too high. 

 

2 hours ago, Conor said:

 

You'll not need the heating past March or so, heating up time of the buffer isn't an issue.

Thanks. What about the heating season though? I assume you'll only have the heat time for the buffer that first time if your heating goes on very low during the heating season? 

 

1 hour ago, dpmiller said:

'd suggest if you want to have a thermostat on the buffer, simply putting it in series with the demand signal from the wiring centre might work for your case?

Thanks. I don't specifically want a stat in the buffer, I wad just trying to figure out what I needed. 

 

The only thing I was trying to avoid was the buffer being kept at temp when there is no heat demand from the room stat. I.e. out of heating season. 

 

However I might be worrying about nothing. The energy might be very low for keeping the buffer warm? 

 

If I use a buffer ( to avoid short cycling - even though i might not need it) Do you put a buffer stat and a room stat and then the buffer just turns on the ASHP as and when it needs topped up? What's the norm?  

[SuperJohnG]

6 minutes ago, Nickfromwales said:

Correct. 
 

In the summer your heating will all be in the summer / off position, so I don’t get why your system would come on, to heat anything? 
Summer sees the ASHP > UVC directly, and the buffer remains isolated and dedicated to the heating primary circuit only. 

Crossed over thanks. 

 

That's how I thought. Do you use a direct or indirect buffer? 

[SuperJohnG]

30 minutes ago, Nickfromwales said:

In the summer your heating will all be in the summer / off position, so I don’t get why your system would come on, to heat anything? 

It will be off, so just the ASHP to heat the UVC. Thanks

[SuperJohnG]

@JohnMo I've been doing a lot of reading on here. 

 

I think I am now not going add a buffer, I get it now as we will have one single zone and the volume will be about 70l which is more than the defrost requirement. 

 

Going on 20W/m2 Vaillant are saying a 7kW unit will do and that will provide 136% of the heating requirement (5.14kW at 100%)

 

I was going to get a 10 kW unit which would provide 195% of the heating requirement. However I'm slightly concerned I might be oversizing the unit? It's only for a few days we get down to -10 most of the winter we are around 0-10 degrees during the winter. 

 

Is it detrimental to have an oversized unit? 

 

Now I'm guessing at 20W/m2 but if it was 30W/m2 I'd need 7.8kW of heat input. A 7kW unit might not be big enough? 

 

I'm trying to decide if I need a  7kW unit or a 10kW? So I can order it. 

 

Also a realistic W/m2 value for this house? 

 

Vaillant want to use 45W/m2 (in the absence of an MCS installer doing the heat loss calcs) which I think is way too high. Using Jeremy Harris spreadsheet I get roughly 13.5W/m2 on say a 0 degree day, 16.6W/m2 on a -5 day and if we go extreme to -20 day I need 25W/m2. We are a SIPS build with 50mm PIR internally removing mostly all thermal bridging and our mid build air test was 0.8 ACH. 

 

In reality using the average values for Jan which is the coldest month in my location (west coast of Scotland) we only need 9.4W/m2! 

 

In reality in writing this I have probably come to the conclusion the 7kW unit would be ideal! 

 

Does my maths make sense to everyone? Someone agree so I can order the ASHP and no buffer. 

 

 

[JohnMo]

Are you putting in any secondary heating like a stove, if so that what takes you down to the lower temperature. No way would you size for -20.

 

Valiant are talking out of their a***

 

 

Using MCS guidance you would size for the 99.7 (think that's the right number) likelihood lowest temp.  So the -9 and -10 temps would not be included.  So you are likely to be sizing or -3 or 4.

 

So if you use the 16.6W/m2 fig you will not be far off.  Then you need to allow time for cyl reheating also in your daily allowance then divide by 24. Allow some time for defrosts, although not that likely at the lowest temperature.

 

Most the time your conditions will be warmer than the design case, so an oversized unit will be struggling with turndown.

 

I would be going the smaller unit no buffer and run weather compensation. Or run weather comp, a degree or so hotter than needed to allow your floor to batch charge, with the thermostat tripping the ASHP off at 0.5 degree below your ideal temp.

 

Not sure I agree with the thermostat on buffer as by default you will never run at the lowest temperature possible, so CoP will always take a hit.

[SuperJohnG]

6 hours ago, JohnMo said:

So if you use the 16.6W/m2 fig you will not be far off.  Then you need to allow time for cyl reheating also in your daily allowance then divide by 24. Allow some time for defrosts, although not that likely at the lowest temperature.

Thanks @JohnMo 

Vaillant size based on -3.9 degrees. So the 16.6 figure would be pretty close. 

The cylinder reheat times are the reason I was going to get a 10kW unit. How do you work those out? 

 

We actually have a borehole and 1000l storage inside the thermal envelope, so our water is in reality preheated before it goes in the cylinder. Even if we allow say 15 degrees the water will be preheated to. 

 

We have an electric fire with a 2kW heater which would cover the -20 scenario. 

 

We are going to go no buffer I think, and run weather comp so we can get that highest possible COP.  If we run the floor at say 27 degrees or so with weather comp does the ASHP shoot for a low temp based on the feedback from the room stat? I assume I'll need the vaillant unit? 

Edited April 13, 2023 by SuperJohnG

[JohnMo]

If you are heating the whole contents of a 300l cylinder from 15 to 48, that's about 12kWh.  So from cold to hot about 2 hours (7kW unit).  But that should only happen once the rest of the time you only be heating about half the water or less, so about an hour. Assuming you have a big coil in the cylinder.

 

If you're well insulated your floor temp will only be a couple of degrees warmer than the room. So the floor temp will be around 22-23. With weather comp you set a curve based on outside temperature, the colder it is the warmer the flow, thermostat is only there to help stop overheating so set a couple of degrees hotter than target.  Ours flow temps vary from 25 to 32. On UFH based on our house I can operate in two ways

1. Run WC and let it look after itself, set the thermostat slightly higher than target temp. This works best for us on a gas boiler when consistently cold.

2.  A variation of the above is set WC curve a degree or so higher than needed, use a thermostat (0.1 degree hysteresis) with a combination of time temp targets to force heating on when house temp drops and off again just below target temp. This basically batch charges the floor overnight based on outside weather. The heating is off for about 14 to 16 hours. This works best at start and end of heating season, when there is plenty of solar gain in the day, our heating hasn't come on for about 10 days now.

 

Note we have 300mm loop spacing and 100mm screed, floor response time is slow.

 

Someone will chip in if I'm talking rubbish.

[SteamyTea]

17 hours ago, JohnMo said:

Using MCS guidance you would size for the 99.7 (think that's the right number) likelihood lowest temp.  So the -9 and -10 temps would not be included.  So you are likely to be sizing or -3 or 4.

Is that 99.7‰ (or whatever) temperature distribution or time?

I thought, from years ago when I was involved with MCS, that it was time.

8734 hours out of 8760 hours in a standard year.

Not sure how much real difference it makes knocking off the third standard deviation on the cold tail of the OAT distribution in reality. Probably none, but it does depend on the variance (spread) kurtosis (tailedness) of the distribution, and skew (lopsidedness).

Edited April 14, 2023 by SteamyTea

[JohnMo]

Just had a proper look at the MCS version I have

 

3.0 Sizing the Heat Pump

3.1. General

Getting the design and installation right

Sizing the heat pump correctly is of paramount importance. Various field trials seem to indicate

that accurate sizing is important in order to maintain efficient running of the system

Heat pumps should be selected as closely as possible to the design heat demands.

MCS Standard MIS 3005, requires the unit to achieve 100% of the duty at an external temperature condition exceeded for 99.6% of the year, if

reasonably practicable...

[SteamyTea]

48 minutes ago, JohnMo said:

reasonably practicable

Yes. I think, since 2012/13 they have clarified.

Did it specify the temperature resolution i.e. 0.1⁰C.

Again, in reality not that important.

[JohnMo]

Didn't notice anything on resolution 

[SuperJohnG]

On 14/04/2023 at 08:16, JohnMo said:

If you are heating the whole contents of a 300l cylinder from 15 to 48, that's a out 12kWh.  So from cold to hot about 2 hours (7kW unit).  But that should only happen once the rest of the time you only be heating about half the water or less, so about an hour. 

 

Note we have 300mm loop spacing and 100mm screed

Thanks for this @JohnMo. 

 

We have 250mm in our ring beam and 150mm concrete throughout on 200mm spacing so I think we will be very slow. 

 

The thermostat we have is the Wunda one we bought with the UFH however I think for the WC we need Vaillant unit the Sensocomfort? 

 

Edit: Oh I didn't notice the reply usually I do so apologies for coming back so late