Buffer tank confusion

[Matt540]

I’m struggling to find a calculation or rule that explains if a buffer tank is required or not. 
 

I’ve spent alot of time trying to research but all I seem to find is two groups of people who are either for buffer tanks (reduce cycling etc) who say a buffer is needed in every scenario or the smaller group who say you don’t necessarily need one and you can run a very good system without a buffer. Both sides have valid reasoning behind them but I struggle to find a solid answer how you work out if one is required. 
 

I understand it’s to do with the minimum system volume. I’m looking at a 5kw vaillant system and all the tech specs say regarding volume is 15/40l for thawing mode. 
 

i will be running UFH on the ground floor  as a single zone in one open system with 100mm centres to increase system volume and allow for lower more efficient flow temps. along with DHW. So all I’m trying to understand is how I work out the best system design, buffer or not. 
 

thanks

Matt

[JohnMo]

My understanding is.

 

You need around 10-15l for each kW your boiler can generate.  So if you have a system volume of over 75l you will be ok without a buffer.  The issue come if the have lots of zone or just a couple, where one zone may be calling for heat and it's volume is small.

 

One thing you can include in a buffer is an immersion, so if things go wrong you have a fallback heating system.

 

[Marvin]

Hi @Matt540

The buffer tank is, designed to avoid short cycling. 

 

A cycle begins when the ASHP starts the heating the circulating water, and continues until the return water temperature meets the ASHP "hot enough" set temperature, and then turns off until the water temperature becomes lower than the "turn on" set temperature.

 

The importance of this is that in tests its noted that an ASHP needs to get going before it reaches its best performance (COP) and then when turning off it does the same in reverse.

 

Because of this the ASHP should aim to heat for at least 10min.

 

The problem of Buffer or no Buffer is to do with the quantity of water in each individually controlled emitter zone.

 

We have a buffer and radiators which have TRVs ( TVRs because we get north south solar gain problems).

 

If we didn't have a buffer tank, and only one radiator was to work, the small amount of water in the system (ASHP to one radiator and back) would heat up quite quickly causing the ASHP to short cycle thus being inefficient.

 

So to answer your question in my opinion the reason some use buffers and some don't is dependant on what the ASHP is connected to.

 

It appears that radiators with no TVRs may work and large areas of UFH may work.

 

However as @JohnMo says buffers usually have the option of an immersion to give additional heat in exceptional circumstances like if the ASHP breaks down (assuming like me you have a pump circulating the water from the buffer to the emitters) or a very cold spell.

 

Hope this helps

 

Marvin

 

 

Edited March 30, 2022 by Marvin
Clarification

[Matt540]

Thanks @JohnMo and @Marvin

 

understood and all makes sense.
 

im just thinking out loud and may be completely wrong but what I don’t understand is, if I have a single zone under floor heating with flow set to achieve deltaT of 5° then the return to the ASHP will always be 5°c less than the output? So no cycling? 
 

the only “cycling” then could be the room stat demanding the heat on/off?

 

if that happens then surely you can reduce the flow temperature of the ASHP to the point where it naturally heats the house to the desired temperature and just ticks over (long cycle time) regardless if you have 30L or 300L in the system? 

[Temp]

I think ASHP have a minimum power output. So you need to be sure the minimum load is higher than that or a) the temperature of the whole system must rise or b) it must cycle..

 

https://mcscertified.com/wp-content/uploads/2020/07/Heat-Pump-Guide.pdf

 

Page 35..

Quote

 

Example: A 12kW nominal inverter controlled heat pump unit (at 7oC ambient and 35oC water flow) can provide an output of 4kW at maximum turn down (minimum output) at an ambient of 12oC and weather compensated flow temperature of 30oC. What is the minimum water content to ensure cycling does not exceed six starts per hour?

 

Assume acceptable temperature drop of fluid is 5oC and the fluid has a SHC of 4 [kJ/oC kg] because it has antifreeze in it (normally 4.18 [kJ/oC kg]) Total heat energy required: 4 [kW = kJ/s x 60/6] [minutes/number] x 60 [secs/min] = 2,400 [kJ] Mass of water required [kg] = 2400 [kJ]/ (4 [kJ/oC kg] x 5 [oC] = 120 [kg] Assume 1 kg of water = 1 litre, therefore 120 litres required.

 

 

[JohnMo]

The theory what you propose, may seem plausible but.

 

If the system contains 30l and you have a flow rate of 6l/s, your heat demand is 3kW, your whole volume of water would have 3kW given to it every 5 seconds, it would very rapidly heat up beyond your control parameters.  The heat pump would short cycle.

 

300l would take 50 seconds for the full volume to flow through the heater.  So a given amount of heat has more to work on and is unlikely to lead to a rapid heating event

 

 

[Marvin]

Hi @Matt540

 

I have learnt yet another thing!

Thanks @Temp and @JohnMo 👍👍👍.

 

I will now use this info to check my system! 

 

 

 

 

[Nickfromwales]

2 hours ago, JohnMo said:

The theory what you propose, may seem plausible but.

 

If the system contains 30l and you have a flow rate of 6l/s, your heat demand is 3kW, your whole volume of water would have 3kW given to it every 5 seconds, it would very rapidly heat up beyond your control parameters.  The heat pump would short cycle.

 

300l would take 50 seconds for the full volume to flow through the heater.  So a given amount of heat has more to work on and is unlikely to lead to a rapid heating event

 

 

A bit like passing your hand quickly or slowly over a lit candle. One you wont feel, the other would burn your hand.

 

I've just installed a similar system for a client, with UFH pipes set out at 120mm c's, and a 7kW ASHP. I fitted a low loss header with a manifold that services the whole ground floor from one central room stat. On the cusp of warranting a buffer, but zero space and the ( potential ) repercussions were not sufficient to warrant the upheaval of integrating one tbh.

Current project sees a very different building with a main 2-storey dwelling, but with a few sporadic projecting single-storey 'wings', which will be 3 sides to the elements, so that one will deffo be having a minimum 200L buffer and probably 3 or 4 room stats to properly manage the differing characteristics of each of these spaces eg to maintain one ambient.

[SteamyTea]

10 hours ago, Matt540 said:

if I have a single zone under floor heating with flow set to achieve deltaT of 5° then the return to the ASHP will always be 5°c less than the output? So no cycling? 
 

the only “cycling” then could be the room stat demanding the heat on/off?

 

if that happens then surely you can reduce the flow temperature of the ASHP to the point where it naturally heats the house to the desired temperature and just ticks over (long cycle time) regardless if you have 30L or 300L in the system? 

If the ΔT is too small, the HP will be running, along will all the circulation pumps/control gear, but little thermal power will come out the system.  Think of it as a car idling outside a school.

Generally, maximum power transfer is when you get half the energy out of something, it is why cooling curves are not straight lines.

Takes two minutes to drop 50°C, then over 10 minutes to drop the next 50°C.

So you may be running your complete system for hours, while it is only delivering a few watts.

A buffer will allow you to set up the HP in its most efficient range, then switch off, then the heating circulation pump, which is not bothered by the temperatures involved, it just pumps, is controlled by the thermostat.  Once the buffer drops to the lowest set temperature point, the HP starts up again, until it reaches the highest set point.

So the buffer temperatures are what turns on the HP, not the room temperatures.

[dpmiller]

^ but with inverter HPs the efficient range is very wide, infact best efficiency will be at much less than full power as you can stay out of the defrost zone for longer.

[SteamyTea]

53 minutes ago, dpmiller said:

^ but with inverter HPs the efficient range is very wide, infact best efficiency will be at much less than full power as you can stay out of the defrost zone for longer.

Yes, the temperature range may be larger, but it still depends on the midpoint target temperature.

So say you want to run the UFH at a mean of 30°C, with a variance of ±2°C. The buffet needs a lower band of greater than 32°C, with a higher band of less than 36°C to allow for mean full energy transfer.

 

Water heating will need a higher overall temperature, and a much larger variation range to keep within the optimum 50% transfer window. Say 40°C to 55°C. 

The lower bound may seem lower than the desired DHW temperature, but it allows for efficient heating from below that temperature.

 

Ideally the ∆T between flow and return will have a mean difference value of ±50% of the desired storage  temperature, and very upwards as the store temperature increases.

Plus a degree of two for losses

[DenkiJidousha]

Is there a good way to estimate or even measure the water volume in an existing UFH system? I can measure total floor area and the external pipe diameter (at the manifolds), but would be a bit of a guess for internal pipe diameter and actual length of each loop. i.e. How tightly they were laid as they snake round each room.

(I am trying to determine if we really would need a buffer for an ASHP retrofit, or if by simplifying the UHF from room based thermostats to single/dual zone we'd be able to avoid it. Avoiding a buffer tank would help with space constraints.)

[JohnMo]

We have 190m2 with a spacing of 300mm (more normal is 150 to 200mm). We have about 400m of pipe in the floor, which approx 46 litres.  A close estimate would be to double my volume of water, divide by 190 then multiply with your area.

 

A 6kW min load of a boiler/ heat pump requires approx 125 litres, for a 10min run time.  Use the calculations in MCS manual I linked to earlier for the calcs.

 

Hope this helps

[DenkiJidousha]

That does help JohnMo - thanks. The house is about 125m2 all UFH, pipe spacing unknown but my guess from a few inconclusive thermal camera images something like you suggest - giving about 60 litres for the house as the effective volume if simplified to a single zone, or only about 30 litres each if upstairs and downstairs are kept as separate zones. So, I think from the examples above we might get away with no buffer if the entire UFH was combined into one zone. But on balance the buffer is probably the sensible choice.