Heat pump running cost - Sanity check

[mk1_man]

Hi all,  my brain is hurting after trying to work out my heat loss as there is so much conflicting data. On the one hand I know that I run our underfloor generally at 35 degrees via gas boiler. We don't run weather compensation, system based on room thermostats / load compensation ?   That said the heating is mostly on in the morning from around 4am where it heats the 80 - 90 mm screed after that for large amounts of the day the boiler is switched off, house stays around 20 - 21 degrees most of the time and even when heating is turned off at night it only drops to around 18.5 - 19.0 degrees or so, our setback is set at 18 and I don't think it has even kicked in.  Its a fairly sophisticated Viessmann boiler and can modulate down to around 10% of its maximum i.e. it is a 19kw system boiler but most of the time it never seems to run over 30% modulation, often much less.

 

House has been insulated to what i think is a high standard and very air tight, mvhr etc however heat loss calculations from companies are showing around 10 - 12kw which conflicts what I feel it is.

 

Assuming it is 10Kw  does this mean I need a 10Kw heat source running 24 hrs a day   so with a COP of 4.0  that would be a consumption of 2.5Kw per hour 24 hours a day? 

 

Thanks

 

[Garald]

Run your own heating loss calculations; multiply by a safety factor. (I chose 2.)

[JohnMo]

A sanity check is your gas usage. Not sure of your DHW usage, but say it take 7kWh for that, each day.

 

So you need your kWh usage for the heating season (your bills is a good place to find that) so, divide by number of days since you switched on the heating, to get an average per day, take away 7 for DHW. That will be the average kWh per day, double that figure will be somewhere near your worst day. Then divide that by 24 to give a kW heat input.

 

Total kWh / days heating = average kWh per day

 

Average - 7 = average heating day

 

Average day x 2 = worst day

 

Worst day / 24 = heat input kW

 

24 minutes ago, mk1_man said:

Assuming it is 10Kw  does this mean I need a 10Kw heat source running 24 hrs a day   so with a COP of 4.0  that would be a consumption of 2.5Kw per hour 24 hours a day? 

On the worst day - most days not that

[SteamyTea]

1 minute ago, mk1_man said:

Assuming it is 10Kw  does this mean I need a 10Kw heat source running 24 hrs a day

Yes, and to add more conflict to it, no.

 

Overall, if the mean temperature difference, say a ΔT 20°C, produces a result of 10 kW (that is the power), then over a day that is 240 kWh (the energy).

 

Now the confusion will arise that the temperatures inside and outside the house will not stay the same i.e. 0°C outside and 20°C inside.

This will produce a picture like this.

 

Now these are real readings for my house.

The mean inside air temperature (IAT) is 18.2°C, the mean outside air temperature (OAT) 4.1°C, the mean ΔT is 14°C (or K, for kelvin in real science).

Just looking at the ΔT, the yellow line, you can see that it starts high at around 15°C, then drops to about 10°C then rises again back, peaking at 16°C.

The IAT, is fairly stable apart from the dip between 1PM and 2PM (probably cooking and opened the window).

The OAT is stable during the first hours of darkness until 8AM at around 2°C, then rises quite fast 8°C, then the day slowly cools back to 2°C.  It was probably a day with a sunny morning and then a slow wind direction change (to the SE) during the afternoon.

 

The relevant part as far as heat pump (or any heat source) part is concerned is the biggest temperature difference (ΔT).  This happened between 10 PM and midnight, though it can happen at any time of day, and on some days can happen multiple times (spring and autumn stormy days, lowish sun angle and scattered showers).

 

You need to know not just the the greatest temperature differences, but also what performance hit your heating system will have when it is cold outside.  This is more important for A2AHP's as they may have a lower rating at low OATs.

It is usual to size a heating system to cover 99% of the loads that can be expected, that leaves 88 hours when the systems will not deliver enough power to keep the IAT at the desired level.  It is usual to use supplementary heating i.e. a fan heat to fill in that gap (or just be tight and have a slightly colder house).

During those few hours, which may or may not be contiguous, the heating system is still supply thermal energy, just not enough to keep the IAT at the desired level.

It is generally cheaper to run a resistance heater for a few hours than fit an oversized heat pump that will take a performance hit because is cycling on and off.  This is why a buffer tank is fitted as it allows a heat pump to deliver energy for a longer time, which generally translates to a better coefficient of performance (CoP) for the ΔT.  Not always ideal, but better than if the system was cycling.

 

In brief, if your house takes 0.5 kW of power for every Δ°C, when you hit the really large ΔTs, and your pump can only deliver 3kW, rather than the stated 8kW, which is what you need, you have to have supplementary heating, or accept that the ideal IAT (say 20°C) will drop by 10°C, so down to 10°C, not very nice, but not fatal.  These will be very rare events as most houses store a fairly high amount of energy in the first few millimetres of internal wall, and temperature loses are not linear, the colder it gets inside, the slower the energy release is from the walls (and other parts of the structure and fittings).  This will, over a few hours, help to keep the temperature reasonable enough (it is why buildings do not instantly change temperature).

 

[MikeSharp01]

2 hours ago, mk1_man said:

However heat loss calculations from companies are showing around 10 - 12kw which conflicts what I feel it is.

This appears to be a very common issue - we are having it as well. The calculations, done with the Passive House Planning Package (PHPP), for our build shows we actually need about 1200W to maintain a dT of 22^oC (K) (-2 OAT to 20 IAT) but we are being told we need an 8kW ASHP by one potential supplier and 6kW by others which to me is madness. However we do need to take into account the DHW needs which will need factoring into the size of the ASHP so I think I need a max size ASHP of 5kW and could quite happily live with 3.5kW.

[SteamyTea]

1 minute ago, MikeSharp01 said:

8Kw

 

1 minute ago, MikeSharp01 said:

6Kw

But oddly

1 minute ago, MikeSharp01 said:

3.5kW

 

[Bramco]

41 minutes ago, SteamyTea said:

It was probably a day with a sunny morning and then a slow wind direction change (to the SE) during the afternoon.

 

I've been graphing the difference between OAT and the wind chill temperature (WCT?) which is often 5C lower at night. This can add 20% to the difference between the inside temperature and outside which isn't taken into account with the heat loss calcs which use OAT.  I'm trying to find a 4 ranges between zero difference and 25C difference for the heating needed. We run the ASHP predominantly at night on the cheap Octopus GO rate so the strategies would be maybe 3 hours of ASHP (night), 6 hours of ASHP (night), 6 hours of the ASHP (night) + a few hours of buffer tank immersion boost, then on really cold windy days add a couple of ASHP hours around lunchtime when there might be some solar to cover the electricity.  Obviously the ranges wouldn't be equal parts of the 25C difference and there might be ways to also use the WCT-OAT degC to manage exactly when the additional heating is needed.

 

Has anyone else looked at the wind chill factor when doing the heat loss analysis?

 

Simon

[SteamyTea]

5 minutes ago, Bramco said:

Has anyone else looked at the wind chill factor when doing the heat loss analysis

As windchill is a function of evaporation, relative humidity and surface boundary windspeed, it will make for an interesting partial differential equation solution.

Many years ago I looked at the correlations between windspeed, wind direction, solar power and air temperature.

Basically as windspeed increased, so did air temperature (for my location), solar power went up with more northerly winds, but I did not look at relative humidity.

I am not sure how buildings are affected by external evaporation.  One way to check would be to put calibrated probes just under a sample of the wall surface and another on a similar sized bit of glass, or an identical bit of wall surface but sealed to stop evaporation i.e. a clear varnish. The idea being that glass does not absorb water and will dry off fast, while brick or render will absorb a bit and evaporate over time.

If it is done at night, then solar power does not enter the calculations, daylight testing would become a secondary experiment.

A more comprehensive experiment could be done with added sensing i.e. RH, solar power and airspeed near the samples.

[MikeSharp01]

1 hour ago, SteamyTea said:

1 hour ago, MikeSharp01 said:

8Kw

 

1 hour ago, MikeSharp01 said:

6Kw

But oddly

1 hour ago, MikeSharp01 said:

3.5kW

 

Surely not - but your observation is valuable and has resulted in my editing my post to better reflect the reality of units now. 

[SteamyTea]

2 minutes ago, MikeSharp01 said:

has resulted in my editing my post to better reflect the reality of units now. 

Good, I can breathe properly again, and hopefully the mouth foaming will stop.

[mk1_man]

Hi guys,  my last 60 days gas consumption averages 1m3 for hot water and 6.5m3 for heating - Viessman boiler gives gas consumption for heating and water separately which is handy also verified against meter readings.

 

So if my logic is correct  6.5m3 x 1.02264 (correction factor) x 40 (calorific value) / 3.6 (kWh conversion factor) = 74 Kwh per 24 hour period.  Divide that by 24 = 3.08 Kwh      Appreciate my boiler is not on 24 hours a day but to keep house warm at 21 degrees if it was on trickle heating this would be true?   If that is the case then how have two MCS companies say my heat loss is between 10 & 12 Kw ?

 

is my maths / logic correct?  My own calculations marry up with my gut feeling that a 7kw Vaillant should be more than sufficient.

[SteamyTea]

23 minutes ago, mk1_man said:

74 Kwh per 24 hour period.  Divide that by 24 = 3.08 Kwh

The h cancel out, so becomes kW.

Not Kw.

[JohnMo]

2 hours ago, mk1_man said:

Hi guys,  my last 60 days gas consumption averages 1m3 for hot water and 6.5m3 for heating - Viessman boiler gives gas consumption for heating and water separately which is handy also verified against meter readings.

 

So if my logic is correct  6.5m3 x 1.02264 (correction factor) x 40 (calorific value) / 3.6 (kWh conversion factor) = 74 Kwh per 24 hour period.  Divide that by 24 = 3.08 Kwh      Appreciate my boiler is not on 24 hours a day but to keep house warm at 21 degrees if it was on trickle heating this would be true?   If that is the case then how have two MCS companies say my heat loss is between 10 & 12 Kw ?

 

is my maths / logic correct?  My own calculations marry up with my gut feeling that a 7kw Vaillant should be more than sufficient.

You need to add the allowance for DHW heating also to the 3.08kW.  Plus the 3.08kW is an average figure it doesn't account for the worst heat day or the design case, of around -3. So DHW will add 0.5kW and doubling up the consumption to allow for the coldest days gets you to, 6.7kW ball park.

 

Now the bit missed by a lot of people, is every manufacture uses a different scale to size the marketing size of the heat pump. So you need to look at the technical datasheet for each specific heat pump and look at the max output at the coldest design day temp. Based on the above an 8kW at -3 het pump is close to what you need, but that's only a ball park, you real heat loss calculation to size correctly

 

 

[mk1_man]

Why would you need to add hot water, surely the pump is either heating water or underfloor but not both at the same time?   Also how many days of the winter is it -3, I know its a little risky but I would rather either wear a jumper or add a fan heater for a few days of the year as opposed to being oversized for the majority of the time ?    The benefit I guess of being forced down the route of self purchase and install is that I could also keep the gas boiler as a reserve, easy enough to plumb this way with a couple of valves or if I keep my small LLH I can have both plumbed in parallel 

 

[JohnMo]

29 minutes ago, mk1_man said:

Why would you need to add hot water, surely the pump is either heating water or underfloor but not both at the same time?

Because there is only 24 hours in a day. If you are heating DHW for say 3 hrs that leaves 21 hours.  So instead of dividing by 24 you really need to divide by 21.

 

Undersized is really as bad as oversized, especially if it starts trying to activate immersion heaters to help it cope.

 

Quite like a Hybrid heat pump, smallest heat pump topped up by gas when you get the frequent defrost temp and below (about 3 degs and below).

[Lofty718]

@mk1_man are you using 6m3 of gas per day for heating? this seems quite a lot.. how big is your house?

[SteamyTea]

50 minutes ago, Lofty718 said:

6m3 of gas per day for heating

About 65 kWh I think.

[mk1_man]

just checked again and it is showing as 4.97 m3 / day which I think works out as 56.5 kWh?   That then equates to 56.5 /24 = 2.35 kW each hour to heat house.   Does that mean heat loss is around 3Kw per hour?   I am getting so confused with all of the data being thrown at me recently !

 

House is 250m2

[SteamyTea]

17 minutes ago, mk1_man said:

3Kw per hour

No, on two fronts.

 

kW, is power, Kw is nonsense.

 

Now a W is a unit of energy per second, J/s or J.s^-1. The SI unit for energy is the joule, not Joule, that was the man it is named after.  it is a derived unit made up from SI base units of kg⋅m²⋅s⁻²

 

An hour is 3,600 seconds.

 

So, even correcting the Kw to kW, we have, in your example:

 

3,000 [J.s^-1] / 3600 [s] = 0.83 J, which works out as 2.3 x 10^-7 kWh, so small to be meaningless.

 

Now if your energy usage, over 24 hours, is 56.5 [kWh] then to convert it to power [kW], just divide by the time [24h].

 

56.5 [kWh] / 24 [h] = 2.35 kW (the time, h, cancels out).

 

Edited March 14 by SteamyTea