When is a 50c DHW tank really 50c?

Hardware Heatpump
1

Ok, it’s not as simple as that… some days we may only use part of our Altherma 250l tank, others more so.

Below is the last 3 days. Schedule heats the tank at 1am for as long as it takes, no reheat.

Starting on the left, Tuesday morning, the tank was 33ish, just under an hour to bring it back to 50c and 7kWh of produced heat.
As it was mild, and space heating wasn’t running, I manually ran a hot water cycle that same evening. Another just shy of an hour run and around 5kWh of heat produced.

image
image4058×1750 247 KB

The next day, Wednesday, the hot water seemed to run out earlier than usual.
Thursday morning a full reheat from cold to 50c, about 90 minutes and 11kWh produced heat.

We’ve had three showers today and you can just make out their blips in the line for Thursday. Barely any impact on tank temperature.
Yesterday it was three showers as well, but there just didn’t seem the same amount of hot water to go around.

I don’t have any evidence to suggest this but it feels like when the tank reports, say, 45c, everything below the sensor point (a good 2/3 of the tank as I understand it) could be 15c or whatever the incoming cold water temp is. During the 1am reheat, it doesn’t take long for the warmer water around the sensor to reach 50c leaving everything below it some way below that point. Result is a tank which didn’t fully heat, presumably, since the hot water doesn’t go as far.

If the sensor temp is much lower before the 1am reheat, like 30c or something, we tend to get a full tank for the next day.

Does anyone else recognise this sort of thing?

2

This is probably it. Tank temperature is only accurate at the position it’s measured at, with water above usually hotter, and water below cooler. You can work out how hot the tank really was by how much energy was put into it to reheat it.

How is the tank heated? At what position within the tank is it heated from?

Moving the sensor to another position, or adding a second sensor, may give better understanding as to how much hot water is in the tank. Or get a Mixergy tank.

3

It’s heat solely via the coil using the compressor, presumably the coil is at the bottom half of the tank.

I don’t think the sensor can be moved since it’s inside somewhere, right?
The tank is a 250L (EKHWSU250D3V3)

4

Tanks usually have sensor pockets on the outside, they don’t usually sit in the water itself, there’s a small metal recess that is close in diameter to the senor so it can be inserted and read the temperature of the water that the pocket is within. Some tanks have more than 1 pocket. Mine has 2, 1 just above the top of the coil, and 1 half way down the coil.
I recommend using thermal paste on the sensor to increase the contact surface area between the sensor and the pocket to increase the accuracy, then put some insulation over it on the outside.

1 Like
5

I put these observations down to changes in the temperature gradient (vertically) in the tank, coupled with the fact that the sensor is only at one point.

If we have a cold DHW tank and heat it for a bath JUST before wanting the water, a setpoint of 33C gives plenty of water for a deep hot bath. Leave it for several hours and you get a tepid bath. I judge that while heat is supplied, the gradient is steep but becomes less so once left to stand (even without drawing water off, which must have its own effect). So a reheat in a reduced-gradient scenario may reach setpoint with less input and without the top water being as warm as the first heat..

With the 33C setpoint, the tank sensor rises to over 36C about 15m AFTER the HP stops providing input. Drawing water off up to 50m after input stops, the water temp leaving the DHW tank is over 42C.

Data from a couple of pockets + draw-off temp would be revealing…

1 Like
6

This is how top and center/lower third sensors compare for me. A reheat is triggered by the center sensor. You can see how the difference between them is not just a constant offset it changes rapidly with how much water is drawn. It’s totally reasonable to get a hot bath out of a freshly heated tank at target 33°C, as the top might comfortably sit at 40+°C.

17687599389354200047249116806372
176875993893542000472491168063721075×613 53.5 KB

2 Likes
7

I’ve got a few pockets…

image
image1119×569 127 KB

The dips are as DHW is drawn or as the UFH is running.

I played with trying to get a heat map in Home Assistant. I got this

image
image497×619 18.1 KB

[edit]
This is a boiler fed system. I control how hot the tank gets depending on time of day (with HA). For showers in the morning I’ll run the boiler longer to fill the tank more. When the UFH is running, if it needs a top up it does so, but fills the tank less.

The UFH is blended at the tank so it draws off from various points on the tank to get a 35°C feed (which is then blended again at the UFH manifold)

6 Likes
8

I have a Daikin Altherma with a 230 litre tank integrated with the pump, filter etc. If I look on the user interface - Sensors menu you can see two hot water temperatures at the upper and lower parts of the tank. Is it not the same on yours?

9

Sadly no, I get just ‘Tank temperature’.

Love that @borpin !

1 Like
10

The tank will have a thermocline level. When the tank is re-heated, almost the whole tank will be at the set-point. Then as hot water is drawn-off the top, cold water enters at the bottom the thermocline level will move up the tank as the hot water is used. In my experience, the thermocline is often quite abrupt, you can think of it like a horizontal line, but overtime the hot and cold water will eventually mix, and the layer will become more diffuse.

Having the heat pump temperature sensor low in the tank is a bad idea, this will give the false impression that the DHW tank is cold after only a small amount of hot water has been used, when in fact 80% of the tank above the sensor is still hot. This will result in the heat pump attempting to re-heat the tank when most of it is still hot, this is not optimal for efficiency.

3 Likes