Overnight temp drop 0.3/hour in my MVHR+SIP house, is this good?

[SteamyTea]

3 minutes ago, Radian said:

But in the winter there's no sun on those faces

There is on some of them.

Winter generation ideally needs South Facing and around 15° from vertical.

Was just a thought, I quite like the look of hexagon and pentagon, housing as it a little different.

Would be harder to make, and if an architect was involved, crap to live in.

[Thorfun]

if i may join in here....our house isn't finished yet and isn't airtight (we have 2 openings to the garage and the garage doors are sheets of OSB on hinges (i have a piece of PIR leaning against a wall between the house and those openings). we only have heating in the basement at the moment. yesterday at about 4pm i turned off the heating as it's dry down there now and it's just using electricity for no real reason. when i left the build it was 19.6°C in the basement and 14.8°C on the ground floor. overnight temps were -3°C.

 

this morning when i went and checked (at about 10am), the basement was at 16.8°C and the ground floor at 12.2°C. i'm pretty happy with those loses for now considering we're not even airtight and it's fricking freezing outside (Mr. Bigglesworth). so it really makes me feel good that eventually we should have an efficient and warm house with minimal heating loses. when i finally get it finished that is!

[SteamyTea]

8 minutes ago, Thorfun said:

if i may join in here

So around -0.15°C/h

 

Pretty good for an unfinished place.

[Thorfun]

2 hours ago, SteamyTea said:

So around -0.15°C/h

 

Pretty good for an unfinished place.

yeah. obviously that's an approximation but i'm still very happy. all those hours, days, weeks, months fitting insulation, AVCL and taping will hopefully be worth it in the end!

[Blooda]

Not really looked at this before, but just downloaded the two Elitech loggers, one in the lounge one in the Kitchen Diner.  And besides whats it telling me there seems to be another load of questions.

 

House = self build finished last year, insulation better than building regs air tightness around 3.5 - not as good as we wanted but COVID prevented constant survalence and monitoring.  ASHP and UFH. 

 

 

Lounge - We had the Log Burner Going,

Kitchen Diner - Residual heat over 21 from Cooking etc.

 

Blue Arrow = heating off Set Back time @ 18 degree

Orange Arrow - 11pm to 6pm Trial time

Outside temp around -2

 

Kitchen Diner  lounge 17-01-2023 23:02:23 21.5 23.8 18-01-2023 06:02:23 20.4 20.1       Drop in 7 Hours  1.1 3.7 Drop Per Hour  0.157 0.529

 

 

 

 

 

Edited January 18, 2023 by Blooda

[Jenki]

I'd settle for that👍. 

Static caravan life:  went to bed last night at 23:00 lounge temp 19.2°C. Outside temp -1°C (-5.2°C wind chill)

this am. 05.00 Outside temp 1.2°C (-3.9°C wind chill) Lounge temp 7.5°C 🥶

 

so circa -1.9°C/h  although its more like -10°C hour 1 then  -0.34°C/h  😂

 

I don't have the temperature in the bedroom as that would be scary

 

[ProDave]

2 hours ago, Jenki said:

Static caravan life:  went to bed last night at 23:00 lounge temp 19.2°C. Outside temp -1°C (-5.2°C wind chill)

this am. 05.00 Outside temp 1.2°C (-3.9°C wind chill) Lounge temp 7.5°C 🥶

The best thing we did with our static was install a multi fuel stove, it burned wood in the day and coal over night.  We lit it in November and it hardly went out until March.

 

We survived the winter with the "Beast From the East" in that.  Only had 1 frozen pipe, where the mice had eaten about 12" of pipe insulation leaving a bare pipe.  Crawling under the 'van with a hairdryer and extension lead during a blizzard.

[Jenki]

1 minute ago, ProDave said:

The best thing we did with our static was install a multi fuel stove, it burned wood in the day and coal over night.  We lit it in November and it hardly went out until March.

 

We survived the winter with the "Beast From the East" in that.  Only had 1 frozen pipe, where the mice had eaten about 12" of pipe insulation leaving a bare pipe.  Crawling under the 'van with a hairdryer and extension lead during a blizzard.

Looked at this, and opted for the AIR Heat pump, and I'm still happy with this.  biggest mistake was single glazing 😥.. The stove had too many negatives, no free wood, cost,  dogs etc..  just concentrates the mind to not be in here too long.

[Radian]

9 hours ago, Blooda said:

Not really looked at this before, but just downloaded the two Elitech loggers, one in the lounge one in the Kitchen Diner.  And besides whats it telling me there seems to be another load of questions.

Like why the losses from the lounge are significantly more than from the kitchen?

[SteamyTea]

9 hours ago, Blooda said:

whats it telling me there seems to be another load of questions.

Yes.

Why it is such fun.

 

The trick is to ask a question that can be answered, and tested.

Edited January 19, 2023 by SteamyTea

[Blooda]

56 minutes ago, Radian said:

Like why the losses from the lounge are significantly more than from the kitchen?

I am thinking that it might be the big black, heat absorbing lump of iron, with a hole up to the roof. 🔥  Even though the stove is air fed from outside, the air baffles will probably not fully shut for safety reasons, when it is not on.  

 

Last night no fire:-

The peak in the graph with the star is when I moved them to down load them. 

 

 

Lounge = double heat loss.

 

 

53 minutes ago, SteamyTea said:

Yes.

Why it is such fun.

 

The trick is to ask a question that can be answered, and tested.

Back to first principles. Check the data source.   The two Loggers are now side by side, and will spend 24 hours in the Kitchen and then 24hours in the lounge to see if they are reading the same.

 

 

 

 

 

 

Edited January 19, 2023 by Blooda

[Radian]

4 hours ago, Blooda said:

Lounge = double heat loss.

 

Not quite. Recall the exponential nature of cooling. Both kitchen and lounge need to be at the same starting temperature WRT outdoor temperature. Notice how your lounge was even hotter to begin with in your first test and the loss you recorded was higher.

[SteamyTea]

3 minutes ago, Radian said:

Not quite. Recall the exponential nature of cooling. Both kitchen and lounge need to be at the same starting temperature WRT outdoor temperature. Notice how your lounge was even hotter to begin with in your first test and the loss you recorded was higher.

May be loosing twice the amount of energy though.

Temperature is not heat (the old word for energy).

 

But yes, it is really only worth comparing within the same temperature ranges, and then the energy losses will be the same, one the differences in form have to accounted for.

 

A wood burner will (expletive deleted) it up though, half the time it is so hot the energy is going up the chimney, the other half of the time, it is cold, and draws air back down the chimney.

[SteamyTea]

4 hours ago, Blooda said:

The peak in the graph with the star is when I moved them to down load them. 

They show lovely cooling curves.

May just be an offset.

You can use anomalies.  Take the mean and subtracted the variables.

 

Was the kitchen extractor left on.

 

 

Edited January 19, 2023 by SteamyTea

[Blooda]

1 hour ago, SteamyTea said:

They show lovely cooling curves.

May just be an offset.

You can use anomalies.  Take the mean and subtracted the variables.

 

Was the kitchen extractor left on.

 

 

 

Umm .  We have MVHR, it drops down at 10pm til 7.30 to around 20%.  And thinking about it now the logger could have been placed to near one of the vents.  

 

A lot to start to look at.  and maybe need some more logging points.   What does anyone else use for logging temps ?  I currently have a couple of these:- USB Temperature Data Logger, USB Temperature Monitor Manufacturer (elitechlog.com)  can't read them remotely and they warm up when I plug them into the laptop.

 

 

 

 

 

 

[Radian]

3 hours ago, Blooda said:

What does anyone else use for logging temps ?

A whole bunch of these 

Used to be under a tenner. Maybe they still are on ebay. There's good hack that makes them into BT beacons then you can catch the beacons with something like a Pi or ESP32. I went for the latter. It listens for a bit, identifies the beacons (amazingly good range) and sends the readings out over MQTT. A flow on Node Red listens to the MQTT and sends the data off to a MySQL database.

[SteamyTea]

2 hours ago, Radian said:

sends the readings out over MQTT. A flow on Node Red listens to the MQTT and sends the data off to a MySQL database.

You are going to have to post up a tutorial on this one day.

 

I intend to work out how to make an ESP2866 (only because I have some) go to sleep until the LED on my meter flashes, then send the data to a hard drive somewhere probably on my TOR webserver, which is working incredibly reliably so far.

Triple encrypted all the way, and no need to pay anyone for web hosting.

[Radian]

8 minutes ago, SteamyTea said:

You are going to have to post up a tutorial on this one day.

 

That's true. The link to the Hackaday site nicely covers the adaptation of the sensor to free it from its cloud and send out regular temperature & humidity BLE beacons.

 

On the other hand, grabbing the beacons using an ESP32 and sending the sensor data over MQTT isn't something I've seen published so I cobbled together some code of my own - which is a bit of a problem as it's got hard-coded parameters that would need to be adjusted to suit the local WiFi and MAC address of the beacons. Not a show-stopper I know, but it could have been so much tidier and better written! Also, the other slight embarrassment is that I've never had the time to get the hang of GIT for sharing code. Happy to share the C source though.

[TerryE]

Just another data point for @SteamyTea's table. ATM, the external temp is 0.3°C and falling and our daily ripple is about 1.3°C and we only heat the house in the off-peak window (midnight - 7AM).  The temp peak varies slightly across the various rooms but a typical value in this cold spell is at around 11AM and the minimum at around 2AM, so the house drops a little under 0.1°C/hr.   When the average outside temp was 6°C, the daily ripple was just under 1°C.

[NSS]

On 17/01/2023 at 12:38, SteamyTea said:

I have found that with my house, a small terraced place, that the major influence is temperature difference.  Wind direction and Solar Gain have very little effect.  This is not surprising really as when it is windy, it is predominately from the South West, which is warm, but with cloudy skies.

 

Looking at my data for the cold spell last year, 27/11/2022 to 17/12/2022 and grouping the means by hour of day, this is what happens.

 

 

Then looking at all the negative slopes I get this.  Added a linear trend line as it makes the arithmetic easy.

So Midnight to end of 4 AM the temperature drop is -0.180°C.h^-1 

 

9 AM to end of 11 AM, slope is -0.119°C.h^-1 

 

 

And finally 3 PM to the end of 11 PM, the slop is -0.104°C.h^-1 

 

 

This gives a mean temperature drop of -0.134°C.h^-1 

 

This is only part of the story though.  As I know that the major influence is eternal temperature, I can plot and calculate the slopes against temperature difference by hour.

This again shows three periods when the temperature slope is negative, with the final one being only two hours.

Note that the y-axis has changed from temperature to temperature difference.

 

 

So Midnight to end of 4 AM the temperature drop is -0.180°C.h^-1 

 

 

 

8AM to end of 12 PM the temperature drop is -0.477°C.h^-1 

 

 

And finally 10PM to end of 11PM, slope -0.227°C.h^-1 

 

 

That is a mean difference of -0.296°C.h^-1 

 

 

I am not sure which is the most important number -0.132 or -0.296°C.h^-1 as it really depend on the internal temperature range, which has to be within the comfort range.

I aim for a mean room temperature of 20°C, but during this cold snap, the mean temperature dropped half a degree to 19.51°C, with a range of 2.05° (min 18.37°C, I can live with that).

The external temperature, from the local WeatherUnderground station (reports higher than the new probe I have put in by ~1.1°C) is a mean of 5.86°C, a range of 2.07°C and a min of 5.20°C.

 

This can be plotted two ways again, x-axis as external temperature or as external temperature difference.

 

 

 

The slope changes with a value of 0.635°C increase in room temperature for every degree rise in external temperature, or 0.409°C rise for every degree rise in external temperature, the inverse is also true, so the same drops in internal temperature for every degree drop on external temperature.

 

Not sure if any of that helps any, just pointing out that thermodynamics is never as simple as we first think.

 

 

 

Just noticed I did not change the times on the third chart down from the top.  Is is 3PM to end of 11PM, and I also have a typo that makes it hoours, which is just me pretending to be French.

And to think some of my friends/family think I'm a bit if a nerd 🤔 

Edited January 20, 2023 by Nickfromwales
Assuming you meant nerd? :D

[SteamyTea]

7 hours ago, NSS said:

friends/family

That is the difference between us.

[S2D2]

On 18/01/2023 at 09:20, Radian said:

 

The cold hard reality of the seemingly innocuous 0.3 loss is driven home by how much energy is consumed to make up those losses. I plotted a selection of average daily deltas against the day's gas consumption for the house before and after getting CWI installed at the end of 2022. 

 

WIth gas at 11p/kWh some days were costing us £20 during December even with the CWI. Winter 2021 was a lot milder hence the lack of data points at the top end. The deviations from the trend lines reflect the days when DHW was used more or less.

 

At least the CWI shows up a bit.

In my opinion this is the most reliable method for comparison, ours came out at something like 4.1 kWh/hdd. Remember to include the electricity consumption which is lost into the house as heat too. Adjust if you have any CoP input. Easy enough to factor in exposed perimeter giving an average U value which seems to me the best metric for direct comparison. You don't even need to plot it, just take your annual usage and divide by the total HDD for the year, see: https://www.degreedays.net/. Plotting daily does allow you to filter out the summer days which will skew the regression fit quite significantly once you factor in electric consumption.

 

Mine came out at 0.66W/m²K if I remember rightly, plenty room for improvement but it is a 90's house.

[Blooda]

19 hours ago, Radian said:

A whole bunch of these 

Used to be under a tenner. Maybe they still are on ebay. There's good hack that makes them into BT beacons then you can catch the beacons with something like a Pi or ESP32. I went for the latter. It listens for a bit, identifies the beacons (amazingly good range) and sends the readings out over MQTT. A flow on Node Red listens to the MQTT and sends the data off to a MySQL database.

 

 

Some nice features

• Special mode for babies: baby mode. When the temperature or humidity exceeds the comfort of the baby, Mijia app automatically reminds you of suggested temperatures and gives you the opportunity to give you and your family a smarter guardian
• Very accurate: Built-in Swiss feeling, high precision to not miss the temperature of 0.1°C and humidity of 1% relative humidity, to remind you and your family in time

[TerryE]

18 hours ago, SteamyTea said:

I intend to work out how to make an ESP2866 (only because I have some) go to sleep until the LED on my meter flashes, then send the data to a hard drive somewhere probably on my TOR webserver, which is working incredibly reliably so far.

Nick, the easiest way is not to bother.  If you've got the Wemos or demo boards, then these can be powered through a miniUSB so any old USB plug can be used.  When in light sleep, they only use 10s of mA so the cost is trivial.  Not sure if a cheap powerbank will work with that trickle level, but that could be another alternative.  Deep sleep in effect powers down everything apart from the RTC RAM, so you can still put a pulse counter in this and use deep sleep.  The trick is to push the count using MQTT or equiv every 10 min or 0.1 mWh, say, and that way you only need to wake up Wifi and connect to your SSID on an occasional basis.  As well as the Arduino Dev Framework, you could also use MicroPython or NodeMCU Lua to script this in a HLL.

[Radian]

2 hours ago, Blooda said:

 

 

Some nice features

• Special mode for babies: baby mode. When the temperature or humidity exceeds the comfort of the baby, Mijia app automatically reminds you of suggested temperatures and gives you the opportunity to give you and your family a smarter guardian
• Very accurate: Built-in Swiss feeling, high precision to not miss the temperature of 0.1°C and humidity of 1% relative humidity, to remind you and your family in time

 

Which is why it's worth hacking it clean out of the cloud! It seems the dude that cracked it spent some considerable time reverse engineering the firmware and setting up the web based flasher for his custom version.