Heating cylinder via heat pump

[JohnMo]

Been doing some experiments which are still ongoing. During the tests below cloudy sky and PV generation 0.5kW throughout.

 

Heating a 160L cylinder with ASHP from 42 to 45, I seemed to use around 2kWh of electric to drive the heat pump from main import meter. The heat pump is 6kW. The theoretical energy input required in the cylinder is 0.6kWh for that temp rise.  So on the face of it doesn't look anywhere near efficient. I my be getting good efficiency at the heat pump end making hot water but the conversion to hot water in the cylinder not so good.

 

Experiment 2

Boost cylinder immersion on for 15mins.

Temp rise recorded from 45 to 48 for that period. Electric meter didn't change reading. It's a 3kW immersion, so should have used 0.75kWh.

 

Immersion seems to give a close approximate to energy in verses energy out.  The heat pump seems to loose out in its conversation when transferring heat to cylinder.

 

What are other people finding with their systems?

 

 

[ReedRichards]

Using your immersion heater for 15 minutes you believe it should have used 0.75 kWh but you calculate that the 3 degree change you achieved requires 0.6 kWh.  So is your immersion heater 0.6/0.75 = 80% efficient?  Or is the problem that you can only measure temperature to the nearest degree, so working with a 3 degree change does not achieve good accuracy?    

[Dave C]

Agreed. So many potential sources of inaccuracies here!

 

If the electric meter didn't change reading when the immersion was on, then does that mean it only 'ticks over' when a whole number of kWh is used? In which case your heat pump used perhaps 1.01-2.99kWh of electricity, you can't be sure exactly how much.

 

If the temperature sensor measures to the nearest degree, a 3 degree increase could be anywhere from just over 2 to just under 4 degree increase in reality.

 

Except it's worse - the hot water tank is almost never the same temperature throughout. The water next to the sensor might have increased by 2 degrees, but that doesn't tell you how much energy has gone into the tank, because you haven't raised the entire tank's temperature by the same amount.

 

If you have the entire tank at room temperature, then let the heat pump heat it up to full target temperature, you'll have a better estimation of how much energy has gone in, because at least then you know the starting temperature of the whole volume, and your rounded electricity usage will be off by a smaller percentage. Still won't know exactly - depending on where the sensor is, if it reads 45 degrees, might be a bit of the tank below the sensor at a lower temperature, and possibly the majority of the tank at over 45 degrees.

[SteamyTea]

Repeat the experiment 20 times. Then run a chi square and a student T test on the results.

 

I suspect the it is a combination of turbulence in the cylinder and thermometer/element/thermostat position.

 

[JamesPa]

1 hour ago, JohnMo said:

Heating a 160L cylinder with ASHP from 42 to 45, I seemed to use around 2kWh of electric to drive the heat pump from main import meter. The heat pump is 6kW. The theoretical energy input required in the cylinder is 0.6kWh for that temp rise.  So on the face of it doesn't look anywhere near efficient. I my be getting good efficiency at the heat pump end making hot water but the conversion to hot water in the cylinder not so good.

Focussing on this bit rather than the experiment with the immersion, I would suspect (as others suggest) stratification in the tank and measurement errors.  The hot water will drift upwards from the coil and form a layer at the top and so the sensor lower down might not perceive the rise at all until the whole tank is heated.  As others say meters may be accurate to 1C/1kWh so with such a small change you might reasonably measure something completely different from the fact.  Also, depending on the flow temp and the size of your coil, the heat transfer might be poor (is it a 'Heat Pump cylinder'?) possibly causing excessive cycling.  I would repeat at a lower starting temp for a longer time and also perhaps post some info on coil size and flow temp so a bit of modelling can be done (I do have a crude model of DHW heating with a HP albeit that it assumes the tank is stirred).

 

Of course you may already have ruled all of this out and I might be saying something which you find obvious (in which case I apologise, no offence intended).

Edited May 12, 2023 by JamesPa

[dpmiller]

I've sensors top, middle, and bottom of the tank. The coil and both upper and lower immersions reside below the middle sensor, and this is the sensor for ASHP control. It doesn't respond meaningfully to the convective flow past it, only rising when the HW level is right down on it and the heat pump will then cut out. At this stage the top of the tank will be a few degrees higher, approaching the flow temperature. 

[JohnMo]

So how much energy is your heat source using to heat the cylinder contents up?

[dpmiller]

11 minutes ago, JohnMo said:

So how much energy is your heat source using to heat the cylinder contents up?

 

somewhere around 1.5kWh for a reheat back to 54c

[Radian]

4 hours ago, dpmiller said:

I've sensors top, middle, and bottom of the tank. 

Do you log the data to get a time-series plot? I find the goings-on inside my cylinder are a proper mystery although I only record temperatures at the very top and mid-way of the indirect coil. The top half has an immersion fed by excess PV and the coil in the bottom half is fed by a gas boiler. On days with no PV the bottom can be almost 5^oC hotter than the top and stay that way overnight which surprised me greatly. I thought after several hours the hotter, less dense, water would find its way to the top but it doesn't.

[SteamyTea]

Here is what happens to my 200lt, E7 cylinder.

 

[dpmiller]

8 hours ago, Radian said:

Do you log the data to get a time-series plot?

Nope, I just live here and occasionally look at displays...

 

remembering that my setup is a touch odd (it's a TS, and all flow through the ASHP passes through the coil as a buffer) the bottom sensor- which really *is* at the bottom, below the coil and lower immersion, generally sits between 30 and 40c but on days with good PV will end up at 90c...

[sharpener]

10 hours ago, SteamyTea said:

Here is what happens to my 200lt, E7 cylinder.

 

Where is the element located @SteamyTea ?

 

11 hours ago, Radian said:

The top half has an immersion fed by excess PV and the coil in the bottom half is fed by a gas boiler. On days with no PV the bottom can be almost 5^oC hotter than the top and stay that way overnight which surprised me greatly. I thought after several hours the hotter, less dense, water would find its way to the top but it doesn't.

 

Very strange. Perhaps the energy density (power/surface area) at the coil surface is not high enough to set up meaningful circulating currents. For a traditional immersion heater it is much higher (shorter length, smaller diameter) so probably more effective at this. Does the immersion eventually heat the whole tank, or not @Radian ?

 

All in all it seems to be good evidence that stirring the tank during heating might be helpful in getting the most stored hot water in the least volume.

 

Luckily my 210 indirect OSO tank has both immersion and coil near the bottom so I have never had any reason to question any of this. Surplus PV most days.

[SteamyTea]

8 minutes ago, sharpener said:

Where is the element located

It is a standard E7 cylinder, so almost at the base.

There is a secondary element at the top, but that was not used.

Since I collected that data, I have changed the timings so that rather than heat at the beginning of the E7 window, it heats at the end of it. Mainly did that because for a couple of hours it was loosing heat to the house, and as I like a bath in the morning (decided I actually dislike showers and only use them to get work grub off) I wait till the water stops heating, then draw off a bath full, then they cylinder is at a lower temperature, so less losses.

I really should start to monitor it again, maybe with it's own energy monitor as the kWh trace included all usage, so fridge, kettle (I am a very early riser) washing machine etc.

[Radian]

1 hour ago, sharpener said:

Does the immersion eventually heat the whole tank, or not @Radian ?

 

I wish, but sadly no. The immersion element is exactly half the height of the cylinder which is 1200mm and by the evening of a good sunny day it can be 70+ degrees (HW HI):

 

 

The lower sensor (HW LO) has "no idea" about this and continues to cool off albeit more slowly as per Newton's law of cooling.

 

Coincidentally, I just saw this and thought of @SteamyTea😃

 

[Radian]

Hey, just  out of curiosity, do any heat pump controllers have programmable schedules for different hot water temperatures? I've replaced the dumb electro-mechanical cylinder stat with digital sensors and Raspberry Pi for a bit more control over our gas central heating. Because HW usage in our house follows a fairly predictable pattern, I drop the setpoint in the evening to establish a night-setback temperature rather than have it completely unresponsive. I also have a weekly sterilization program which I know many heat pumps do implement.

[sharpener]

Not that I have seen, they mostly have a schedule where they call for DHW at a presettable temperature, or not.

 

Not sure why you would want a setback rather than no call for DHW, because the norm seems to be DHW priority so (subject sometimes to a programmable timeout limit) you would have no CH while there is a potential HW demand. Don't know if the 3 port valve will revert early to CH if/when the HW demand is satisfied i.e. the cylinder is up to temperature, i would hope so.

 

If you get a simple control interface like the CoolEnergy or Grant (Chofu) have, then you have a dry contact input for the HW call and you can drive this how you want (but still only one programmable temp). If I can find my "no frills" R290 I will drive it something like this with relay logic, external timer(s) and web interface.

 

But with your Raspi you could presumably fool them by also spoofing the cylinder temp sensor readings with a choice of resistors.

 

 

[dpmiller]

CoolEnergy controller has a 4- point scheduler so yes, you can alter DHW setpoint. I do just this to bring the tank up towards the end of the E7 window, ready for the morning showers

[JoeBano]

Chofu controller has 2 water set points