Hello Boycey. Please excuse me if I am misunderstanding your original question and it may be that it is a question of terminology.
My Mitsubishi system uses the Ecodan Hydrobox and I am operating with a weather compensation curve of my own devising through the FTC conrol unit and the result is shown earlier in this thread. This may be different to that which would be appropriate in UK because of the differences in climate here in the Algarve, but it is working reasonably well at this point in the winter.
It takes a little time to experiment with alternative settings, but the point is that you have the ability to design the curve to suit your conditions. I would be happy to advise if you are not familiar with the procedure.
@gary hi, so if your min flow temp is 32c what happens when the weather gets a bit milder, does the house over heat? how do you control the temp in the house, ive been trying to do this "heat geek" thing of remove stats and get the weather curve right
We’re on Vaillant’s 0.55 curve which means 42C at -2C and around 30C at an OAT of 15C.
The pump cycles about once an hour once it’s warmed up the house after set-back. The house upper limit is set at 21C (expanded mode for occasional solar gain) and the rooms are mostly at 20C (bedrooms less).
an update now the weather is much colder and we lost the concept of solar gain months back. We gradually had to increase the curve to 0.65 to compensate for wind chill (3 external walls in the lounge) and we’re on active mode.
The lwt is about 44C now @-2C whereas I misremembered the value for the 0.55 curve we had previously which was more like 40C.
I expect to go back to expanded mode once the February sun breaks through, we get just the occasional warm ray at present.
2kW + Growatt & 4kW +Sunnyboy PV on south-facing roof 9.5kWh Givenergy battery with AC3. MVHR. Vaillant 7kW ASHP (new & still learning it)
@editor Hi its taken me 3 winters to get the curve right for our house, trial and error.
New build well insulated 11.2 kW ecodan
FTC 5
Our weather compensation curve is 41@-4 at the cold end and 33@ 7 at the warm end.
Personally the warm end of the curve was hard to get right.
Currently 2 degrees outside and flow (LWT)is 33 on FTC whole weather curve is -3.
Happily runs at 29⁰ @7⁰ to 11⁰ no cycling it starts to cycle at anything over 12⁰ so I turn it off leave it for a few hours then turn it back on when hall stat starts to drop.
We have taken out 6 room stats and now works on just one stat in the hall. Now hall stat either 22⁰ or 22.5⁰, most rooms downstairs 21⁰ with ufh upstairs rads rooms anything from 18.5⁰ to 20.5⁰.
The hall stat is my marker if it temperature changes I either up the whole curve or minus it depending on if it gets too hot or starts to get chilly.
All stats are set to 30⁰ so always calling for heat, even upstairs mitsibishi wireless stat is set to 30⁰ This also caused system to cycle when we had it set to a target temperature.
Ecodan on all the time no set backs.
Average through Dec 25kWh per day on cold days 1⁰ like today 50kWh we on Octopus intelligent go tarrif.
The Ecodan WC curve doesn’t have presets like you see in other systems it’s fully customisable. So you can set the high and low end at any OAT and
change the gradient as you see fit between those values
Unless I'm misunderstanding something, my FTC6 allows the user to manually set the flow temperature of one OAT between the high and low OATs. This then adjusts the 'gradient' of the lines either side of it (I use the word 'lines' rather than curve, as it is linear, despite looking like curves on the FTC display - more further down). Example:
the 'compensation curve' (as Mitsubishi call it - no mention of 'weather') calculates each degC flow temp as an 'equal' step between the high + low OATs
e.g. using @gary 's values (45C @ -10 and 32C @ 3C), -10 to 3 is 13 steps, 13C / 13 steps = +1C flow temp for every degree C below 3C OAT - i.e. at 2C it's 33C, at 1C, it's 34C, etc.
If you use wider ranges / flow temps, the steps can become non-integer - i.e. 31C flow temp @ 3C OAT would set flow temp to 37.5C @ -3C OAT.
From my findings (I've written a spreadsheet to calculate my curve's flow temps with and without a manual adjustment), it rounds down if the calculated value after the decimal place is < 0.4 (e.g. 37.34 -> 37), it rounds to 0.5 if between 0.4 and 0.5 (e.g. 37.42 -> 37.5), and it rounds up to next integer if > 0.5 (e.g. 37.53 -> 38).
So at -3C OAT, flow temp is 31 + (6 * (14 / 13)) -> 31 + (6 * (1.077)) -> 31 + 6.46 -> 37.46 -> 37.5C
I've noticed I need lower flow temps at/above 1C, so I set my manual adjustment 3C below what the curve would set, which gives a steeper flow temp curve below that OAT.
Which I think is neat, but it would probably be helpful to have at least one more manual adjustment capability, as one is generally limited to using one or two OATs/values around your desired flow temp and this restricts the gradient control. Plus, it has to be an integer value, despite the FTC dealing in 0.5C flow temps :-/
Like bontywood my low end doesn’t go below 32C to prevent cycling and is set at 45C at -10 which gives 37C at -2C
In light of some new sub-zero OATs from the past two days, I'd like to amend what I wrote above (but the forum won't now let me edit the post, despite no replies) - thus:
the 'compensation curve' (as Mitsubishi call it - no mention of 'weather') calculates each degC flow temps mostly* as an 'equal' step between the high + low OATs
e.g. using @gary 's values (45C @ -10 and 32C @ 3C), -10 to 3 is a 13C difference, has 12 flow temps in between, and therefore 13 step ups from 3 to -10 and 14 flow temperatures in total - the difference between each temperature *should be (13C / 13 steps =) +1C flow temp for every degree C below 3C OAT - i.e. at 2C it's 33C, at 1C, it's 34C, etc.
If you use wider ranges / flow temps, the steps can become non-integer - i.e. 0.5C step ups, subject to...
...it rounding - the step value is generally rounded subject to the amount after the decimal place: if it is < 0.38, it rounds down (e.g. 37.34 -> 37), it rounds to 0.5 if between 0.38 and 0.5 (e.g. 37.42 -> 37.5), and it rounds up to next integer if > 0.5 (e.g. 37.53 -> 38)
*however: it doesn't always follow this! and I've not found a 100% rule for how the FTC sets the flow temp, though my calcs are never more than 0.5C away from the FTC values - which I can live with.
it feels like I need to apply a gentle curve correction ratio to get 100% of the values right...
@rhh2348 There are 3 adjustments in the compensation curve menu, the left one adjusts where the colder OAT flow temps start, the middle one adjusts where the warmer OAT flow temp starts.
The one on the right lets you adjust the flow temp within the curve so you can change the gradient and have a shallower and steeper part maybe that will help you with your curve
If you use wider ranges / flow temps, the steps can become non-integer - i.e. 0.5C step ups, subject to...
...it rounding - the step value is generally rounded subject to the amount after the decimal place: if it is < 0.38, it rounds down (e.g. 37.34 -> 37), it rounds to 0.5 if between 0.38 and 0.5 (e.g. 37.42 -> 37.5), and it rounds up to next integer if > 0.5 (e.g. 37.53 -> 38)
*however: it doesn't always follow this! and I've not found a 100% rule for how the FTC sets the flow temp, though my calcs are never more than 0.5C away from the FTC values - which I can live with.
it feels like I need to apply a gentle curve correction ratio to get 100% of the values right...
I don't know if the FTC6 and FTC7 are different in this respect but the FTC7 rounds up and down as you would expect (no odd .38 and .53 rounding points) but the gradient of the line (between the WC points) used to calculate the flow temp is subject to the effects of being stored as a seemingly small(ish) binary number. So the WC "curve" you've set may not be quite be the "curve" used.
Please post your heat pump's weather compensation curves and data as a reference.
Ecodan 8.5 in old farmhouse. 180sq Mtrs.
This cold spell has really given us all a chance to see how our heat pumps respond. We have had our first chance to test the Cold Weather end of our WC curve. Well, the first time since we knew what we were trying to do.
We have been reasonably happy with the results. Although we have noticed that the use of night time setbacks don’t work without some delay of done at -5c.
here is a 24 hour graph showing how the setback put our home behind the target temperature and predictably proved difficult to recover without a +3c add on to the front screen plus/minus display. We were stuck at 18c room temp until we made the +3 adjustment.
Our curve is coincidentally at an exact 45deg slope
We have truncated the water temperature to start at 34c flow temp at 6c ambient (and upwards). It used to be set to 32c at 8c But we were still getting cycling 3 or 4 times an hour so opted to reduce it to 1 stop per hour. This is easily set up on the WC screen. The curve progresses to 40c flow at zero ambient and ends up at 47c flow at-7c ambient.
Our presumption on the setting of the WC curve has been to achieve only the slowest gain of room temperature since, if the room temperature rises quickly then it will basically be inefficient rising past the target temp then switching off - cooling and then ramping up after a few hours. This seesaw heating was our presumption of inefficiency which may or may not be accurate.
To this end we did a recent test. We tried to show the slow recovery after a massive setback (12c room temp) when the house was unoccupied for 8 days. Then to turn on the heating using MELCloud 2 days before we returned home.
the resulting charts were recorded.
It took 48 hours to get up to room temperature. Interesting how the recovery was quicker in the beginning and slowed down as it approached the target temperature.... presumably because of the higher differential between ambient vs internal temperature at the same flow temperature. (Another assumption)
Incidentally we have decided never to use Auto Adaptive to recover from a large setback for fear of upsetting the algorithms of that method of heating which uses the target temp to ramp up recovery. Perhaps other users might have differing opinions of this in using Auto Adaptive heating.
Last point for newcommers - you can move the curser along the base line to see what the flow temperature is at any ambient temperature. You do not need to be in edit mode to do this. In this pic the curser has been moved to -7c ambient and the flow temp can be seen at 47c.
This post was modified 2 days ago 2 times by SUNandAIR
you can move the curser along the base line to see what the flow temperature is at any ambient temperature. You do not need to be in edit mode to do this.
Note that the display only shows the flow temps as whole (integer) numbers but the actual temp set may be different by 0.5C
