Heatpump Optimization

Does anyone has a good algorithm and/or ressources for heat pump optimization?

There might be a lot of parameters to consider:
-) Energy price (including grid costs)
-) HP efficiency based on outside temp. (in the night not so good, but energy costs lower)
-) House heating vs water heating
-) PV production, house consumption
-) Get target temp in time for usage (e.g. shower in the morning)

or might it be not worth do to all this?

Well yes I do but in the end it’s helluva lot of work and complex to handle.
There’s no simple answer to your post so I wouldn’t expect any reasonable answers here.

To get an idea of the algorithm complexity, you might want to check out this and this thread.
It’s about combining a heat pump with a dynamic power tariff. That’s the most promising combo in terms of €€€ (or $$$) and complicated to implement all by itself even without PV and EV.
Adding those obviously adds even a lot more complexity to the game.

FWIW, I’m commercially selling an energy management system based on openHAB to do all of this.
Feel free to attempt coding yourself, but it’s been in development for 3+ years and it requires many, many lines of code to handle that complexity :smirk:
And don’t underestimate how development takes many cycles of testing and optimizing your system.

You can save on average 600€ with hp and ev a year according to a recent study but of course your mileage will vary with your building. There’s also some real world figures in the thread I quoted, amounting to similar amounts.
If that is worth investing a year of work or more now that’s up to you, of course.

1 Like

Hi @robd ,

Are you talking about air-to-air, air-to-water or ground source heat pumps? I assume either air-to-air or air-to-water since the second point on your list was the heat pump efficiency based on outdoor temperature?

The thread that @mstormi was referring to evolved to this solution:

Even though the example in the step-by-step tutorial in the wiki above is about controlling a ground source heat pump, those algorithms can be applied to all kinds of devices as long as you have any kind of binding to control it with openHAB. And it’s not as scary as what @mstormi made it sound like… :slight_smile:

A couple of considerations for starters:

Heating of domestic hot water

  • Heating the domestic hot water is especially during the summer time one of the biggest consumers of energy in many houses. Of course electric vehicles, saunas, hot tubs etc can consume more, but in many households this is consuming a lot of energy.
  • The extent how much you can optimize the heating depends on how big tank you have i.e. how long you can be without heating. For example our house has 180 liter tank in the ground source heat pump and after that, we have a separate 300 liter tank, so total 480 liters. This is enough for 2-3 days consumption and the temperature loss is only ~3 degrees per day. So for our house, we can easily heat it once per day and even skip the heating on one day if the prices are high all day and night.
  • However, if you only have a small tank, then your hands are tied for two reasons. One is the most obvious one, which is that most people don’t like cold showers. The second one is that you need to ensure that the temperature does not drop below a threshold where legionella bacteria likes to live (some countries such as Finland even has a law that requires the temperature to be above the threshold).

Heating of a house

  • There are multiple factors that affect how much you can optimize the heating of a house
  • One of the biggest factors is the question if there is some physical way where you can preserve the heat energy. For example in house, we have 10 cm thick concrete floor and an underfloor water heating. In other words, the ground source heat pump heats the water which circulates in the concrete floor. This, combines with very good insulation means that in our house we can be several hours without heating even during the minus degrees during the winter.
  • A third important aspect is the type of your heat pump. My heat pump is a traditional on/off pump so the compressor is either running or then not. When running with its own automatic logic, the pump uses the concept of degree minutes. When the degree minutes fall below a threshold, the compressor starts and then stops when the degree minutes logic is happy. With this kind of pump, I have optimized the heating so that the heating occurs during 2 or 3 (long) heating periods when the electricity is cheap and the compressor is off on the other times. The water is still circulating also when the compressor is off, but the compressor does not heat it.
  • Not all heat pumps are on/off pumps, there are also so called inverter pumps where the speed of the pump varies based on the heating need. So the on/off logic is not the natural way to approach that one, you want to set the pump to a certain mode when XYZ happen and to another mode when ABC happens.
  • Many heat pumps are SG Ready meaning that they have 4 built-in modes. One is normal mode, one is meant for situation when there is lots of cheap energy available, one is for situations when the prices are up and one is a block for situation when there is shortage of electricity. You can use an openHAB to set the pump to one of these modes when you want, and you can usually configure what the pump does when it is set to one of these modes. For example, you can configure the pump to raise the target temperature when electricity is cheap (or when the heat pump efficiency is good or when you have good solar production). Or you can configure it to lower the target temperature when the prices are high (or there is no PV production / heat pump efficiency is bad).

I’d suggest to go through the wiki of the link above, you’ll get quite a nice understanding by reading those through.

Cheers,
Markus

2 Likes

Thank you for the links!
Bei den Preisen von storm.house brauche ich eignentlich gar nicht mehr daran denken auch nur irgendwas selbst herumzubasteln :wink:

This topic was automatically closed 41 days after the last reply. New replies are no longer allowed.