Sensor Accuracy (Wiser Hub(r))

[SAtewart]

Just thought I'd share some images showing what we all know- that radiator mounted iTRVs overestimate temperature by a few degrees.

This shot shows a working iTRV connected to the vertical radiator and opposite, about seven feet away (on top of the MG sign) about 180cm off the floor level, another iTRV and room stat. All positioned out of direct sunlight and draught-free.

The key for the 48 hour traces below are as follows. Radiator TRV (red), Other TRV (blue), Room stat/display (purple).

The radiator reading is a good 3 degrees above the other two (which themselves show reasonable accordance with each other) when heat is being delivered, but converges to some extent when night time demand drops or goes to zero. I guess that whatever algorithms are applied, eliminating the over-reading of an in situ stat is not reliable.

I can live with this but I really would like to have a temperature offset facility to compensate for these fairly significant deltas between real room temperature and the temperature measured by the controlling TRV.

[SAtewart]

A wee addendum to the above. I shielded the controlling iTRV from the radiator as best I could but the graph trace differences appeared not to have materially changed. So I conclude that the iTRV is not compensating for conduction and convection from the metal valve body into the guts and sensor of the iTRV. You just wonder what the alleged algorithm actually does???

Apart from being able to apply an offset through the app, which is not a function that is available, controlling to desired and accurate comfort temperatures will require setting the scheduled set points with nonsense high values. 

So having said all of this I have no doubt that iTRV's are a vast, vast improvement over the stand alone wax capsule jobs they replaced.

[PhilHornby]

---

@SAtewart wrote:

Apart from being able to apply an offset through the app, which is not a function that is available, controlling to desired and accurate comfort temperatures will require setting the scheduled set points with nonsense high values. 

I'm not at all convinced that any fixed algorithm is going to work satisfactorily. See the graph I posted in this thread.

It shows the temperature from two iTRV's (as massaged by 'the algorithm'), vs the temperature from the Room stat. The Room stat is in an optimal position - the two iTRVs are only about 1.5m apart and yet clearly do not have the same 'view' of the air temperature. (They're next to one another on a wall, but separated by a doorway. The iTrvs are on opposite ends of their radiators - which is why they ended up so close together. The data would make you think they were nowhere near one another. ) 

So having said all of this I have no doubt that iTRV's are a vast, vast improvement over the stand alone wax capsule jobs they replaced.

---

At the very least, they are remotely adjustable.

[njknh]

I upgraded to a Drayton Wiser heating system a few weeks ago and found in my house that the iTRVs were reading high by up to 4 degrees. The compromise solution for me was to use the iTRVs in conjunction with Wiser room thermostats in key rooms and revert to mechanical TRVs in other rooms.

I don't really see how the inaccuracy can be overcome by a temperature offset because the temperature error varies with radiator temperature - what is really needed is to move the temperature sensor further from the radiator(s). Using room thermostats overcomes the problem for me as the Wiser system will take the temperature readings from these instead of the iTRVs in rooms that have both. The problem is that the room thermostats are quite expensive unless you can find them cheaply on ebay etc.