Domoticz

Like to tell you in a few posts about my experience with an all floor heating house with a Nefit Ecomline heater from 1995 and Moduline 400 (was Moduline 3) thermostat and Domoticz.
Recently I modified the Grundfos UPS 25-60-180 mechanical 3 speed adjustment to a Z-wave 3 speed remote control capability. While the conventional mechanical adjustment can still be used. Cost, including Z-wave unit: 130€.

This post contains:
- Objective of this post
- A praise for Blockly
- The struggle between an all floor heating, the Nefit Ecomline and Moduline thermostat
- Why and how of the modification of the Grundfos UPS 25-60-180

Objective of this post.
With this post I like to give my contribution to the Domoticz community.
I'm not into programming anymore (I'm from the age of Algol, Pascal and ADA), so I was looking for an other type of contribution.
I have designed a Z-wave remote control capability for the very common Grundfos pump and like to make this design available for Domoticz fundraising, to stimulate the further development of Blockly.
After a donation of minimal 5€ to Domoticz, see Donation button at http://domoticz.com/ (is that ok Gizmocuz?), I will send, for personal use, the electrical diagram, parts list, mechanical lay-out and 50 setting-to-work steps.
There are other boards where I could publish this idea, like Tweakers or domoticaforum.eu but in this way I hope to create more traffic towards the Domoticz community.

Praise for Blockly
I like to emphasis the importance of Blockly.
For most of us, part of the fun is the making of a domotica system, but for the people around us, it is the use of the domotica system.
An underestimated aspect of Domotica is the maintenance of the system, like slight parameter changes and obsolete hardware replacement.
One day you will sell your house, to whom? To only someone with scripting experience?
My children can adapt Blocky diagrams, the learning curve is very low.

What is missing in Blockly in my opinion are functions like a real IF-THEN-ELSE structure, counters and variable add, subtract, and one-shot timers.
For heating systems Blockly needs to be somewhat more mature.

Many forum members have fun with the use one of the scripting capabilities as Lua, JSON, Python, Bash, PERL, and PHP. They come up with the advice to use their preferred scripting language to solve the missing Blockly elements as mentioned before.

In my opinion Blockly is the key-enabler for the growth of Domoticz!
To develop solutions that can be maintained by others.

Floor heating struggle
When I built this house in 1995 only floor heating was rarely used. The heating installation firm hardly could believe that it was sufficient. But with the introduction at that moment of HR+ glass (K=1.6 W/m2.K) and a good isolation for that time (K=0.4 W/m2.K), my objective was to cash on lower installation cost for the heating equipment. The house is 250m2 and 550m3, while the living space is 210m2 and 480m3. Glass on the ground floor is 26m2 (2/3 south oriented).
Ground floor has 5 groups, first floor 5 plus one for the second floor. One radiator in the bathroom for towels and one on the second floor for very cold days.
The house was delivered during summertime so the heating was marginally tested. Maximum temperature was set to 80°C and 100% power (28kW): the no complaints & customer pays the bill method.
With the shunt, the 80°C supply was mixed down to a maximum of 55°C.
While the installed radiation capability was only 17kW (Nefit power limitation has to be on 6). A floor heating house with a radiator control regime.
The house was always comfortable warm and with a gas bill of around 3000m3, I was not disappointed, although I never understood the cooperation between the Moduline III thermostat and the Nefit modulation.
The floor heating eats all the energy while the Nefit is waiting for increase of the return temperature: a typical radiator control regime. With the ground floor pump on 3, the first floor pump on 1 and the 2 radiators as shunt, the return temperature goes slowly up, and the Nefit establishes a new delta T of around 10 degrees.
Stepwise burning: Versus boost: I tried to do some temperature logging but 19 years ago that was not easy nor affordable.

The replacement of the Moduline III, by a Moduline 400 with more adaptation promises did not help much.

My first persistence temperature measurements started in the beginning of 2014 with 2 WIFI temperature loggers and I saw a strange behavior of heating and waiting in cycles of 10 minutes. I discovered Domoticz and added 16 temperature sensors to 4 FIBARO FGBS001 Universal Binary Sensors to understand Nefit in&out, manifold in&mixed&out and the return temperature of the different groups to log&level the return temperature.
I'm now runnning with a maximum supply temperature of 45°C , maximum return temperature of 35°C, Nefit power limitation on 6=19.9kW. At the start a pump speed of 2 is preferred to fool the Nefit to reach the 45°C faster, while later on a slow speed for optimal heat transmission to the floor is preferred.

Why and how of the modification of the Grundfos UPS 25-60-180
As stated above heating up is optimal with high pump speed. Altough the pipe is in a spiral pattern with low speed it takes an hour before the floor output raises.
When the Nefit is on it's optimal 45°C level, a lower pump speed is prefereable.
The warmwater supply and 25 liter boiler needs to be maintained, by the Nefit, but at the end of the warm-water cycle the residu heat is given to the floor heating system. No problem during winter time, but during summertime that results after 10 start-stops in an unwanted 1°C floor temperature increase.
The pumpswitch has no understanding of speed nor summertime so a regular pomp switch can not handle the formentioned.
Grundfos suggest for floorheating their new Alpha2 L pump: wihout autoadapt and day/night switch. Less energy consumption but no remote control. And still they advice a pumpswitch. The adaptive ALPHA pump seem not so succesfull with floorheating.

This is the start for Domoticz and the Grunfos modification.
The pump has 3 coils that in combination with the capacitor are switched in different configurations of serie and parallel by the mechanical 3 speed handle. My objective was to keep the pump as is and use one switch to make the choice between the original configuration or the remote/local control.
The Z-wave Philio dual switch controls 2 relais. Those create the 4 states, speed: 0, 1, 2 and 3.
For local operation there are 2 push buttons to swich speed: both on is 3, both off is 0. 1 and 2 are the states controlled by one of the buttons.
The local switch disconnects the Grunfos from the relais control and the pump runs on the mechanical choosen speed. Domotica has to be non-intrusive and reversable !
The costs are around 70€ from a well known firm starting with a C and ending with onrad.
The Z-wave Philio is available for less than 60€ and gives with its high RF output power a very good contribution to the Z-wave network.
The remote control unit is operational since last week and I now wil stat with the optimisation of the Blockly control structure. The basic for that is an AN158 switch/meter that senses: on/off, of the Nefic, combined with a Fibaro FGBS001 that senses via a relais the activity: heating or hot water.
For the second floor I made a simple version of the same box with pump on/off and the second switch of the Philio is used with a relais to act as second thermostat directly on the Nefit.

2 more pics of the test setup:

Challenge:
How to control the 2 floor heating pumps

Objective:
When the Nefit heater is ON AND NOT for warm water supply: Switch ON floor pumps
So I check the valve: Heater OR Warm Water Supply (no problem, see further).
To check the "Nefit heater is ON" I have checked the usability of the power reporting function of the Fibaro Wall Plug (FGWPE) and Everspring AN158.
No use is made of the switch functionality (heater must be always on).

Sample and reporting frequency:
"Spot a bear is different from spot a bird"
The agility of the bird asks for more frequent samples (instantaneous observation)
A bear can be followed for its behavior over the time (integral observations)
That's why a power state change must be detected by the instantaneous change, while power cost is defined by the integration of change.

Power reporting:
Both switches do have a Current report (instantaneous, in Watt) and an Accumulated report (integration over the time, per hour in kWh). The Accumulated report is based on extrapolation of Accumulation and the Current usage in case of periods shorter than one hour.
For a state change you need the instantaneous situation (preferable with minimal delay)
For cost you need the accumulated report, because there the duty cycle is taken into account (how many samples within 1 hour showed what amount of energy)

Measurement setup:
I used both, Fibaro and Eversping to measure the Nefit to be able to start the floor heating pumps:
Mains <- Everspring <-Fibaro <- Nefit
Fibaro and Everspring show the following log values when the Nefit is idle:

Fri Jan 23 22:57:37 2015 UI Event triggered: Check_nefit_usage_2
Fri Jan 23 22:57:37 2015 Fibaro FGWPE Nefit: Current Power Consumption
Fri Jan 23 22:57:37 2015 8.9
Fri Jan 23 22:57:37 2015 Fibaro FGWPE Nefit: Accumulated Power Consumption
Fri Jan 23 22:57:37 2015 8.9
Fri Jan 23 22:57:37 2015 Everspring AN158 Nefit: Current Power Consumption
Fri Jan 23 22:57:37 2015 10.0
Fri Jan 23 22:57:37 2015 Everspring AN158 Nefit: Accumulated Power Consumption
Fri Jan 23 22:57:37 2015 10;10200.00
(note: Everspring measures Fibaro + Nefit. Fibaro did not measure it's own usage)

The log suggests 3 times a Float value plus a (concatenated?) String value

Surprisingly only the Everspring AN158 Nefit: Accumulated Power Consumption, can be used in a comparison (the concatenated value!). With log result
Sat Jan 24 16:23:57 2015 11;11000.00
Sat Jan 24 16:23:57 2015 Nefit usage smaller then 20.1
Sat Jan 24 16:23:57 2015 11.0
{note current usage 11W, accumulated usage 11kWh, by coincidence both 11}
Sat Jan 24 16:58:11 2015 83;11000.00
Sat Jan 24 16:58:11 2015 Nefit usage larger then 20.1
Sat Jan 24 16:58:11 2015 83.0

All other values in the comparison give the (infamous) error:
Error: Lua script error (Blockly), Name: Check_nefit_Fibaro_usage_1 => [string "result = 0; weekday = os.date('*t')['wday']; ..."]:1: attempt to compare number with nil
Error: Lua script error (Blockly), Name: Check_nefit_Fibaro_usage_2 => [string "result = 0; weekday = os.date('*t')['wday']; ..."]:1: attempt to compare nil with number

For a better understanding I tried to load the measured values in variables like Float, String, Integer and dummy meter and counter. Comments in {} below

Sat Jan 24 16:23:57 2015 11;11000.00 {current usage 11W, accumulated usage 11kWh}
Sat Jan 24 16:23:57 2015 Nefit usage smaller then 20.1 {comparison works!}
Sat Jan 24 16:23:57 2015 11.0

Sat Jan 24 16:23:57 2015 Fibaro FGWPE Nefit: Current Power Consumption
Sat Jan 24 16:23:57 2015 9.0 {log value is correct}
Sat Jan 24 16:23:57 2015 Fibaro FGWPE as Float
Sat Jan 24 16:23:57 2015 140 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Fibaro FGWPE as String
Sat Jan 24 16:23:57 2015 140 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Fibaro FGWPE Nefit: Accumulated Power Consumption
Sat Jan 24 16:23:57 2015 9.0 {log value is correct}
Sat Jan 24 16:23:57 2015 Fibaro FGWPE as Float
Sat Jan 24 16:23:57 2015 22 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Fibaro FGWPE as String
Sat Jan 24 16:23:57 2015 __ {initial value __ not overwritten}
Sat Jan 24 16:23:57 2015 Everspring AN158 Nefit: Current Power Consumption
Sat Jan 24 16:23:57 2015 11.0 {log value is correct}
Sat Jan 24 16:23:57 2015 Everspring as Float
Sat Jan 24 16:23:57 2015 21 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Everspring as String
Sat Jan 24 16:23:57 2015 __ {initial value __ not overwritten}
Sat Jan 24 16:23:57 2015 Everspring AN158 Nefit: Accumulated Power Consumption
Sat Jan 24 16:23:57 2015 11;11000.00
Sat Jan 24 16:23:57 2015 Everspring as Integer
Sat Jan 24 16:23:57 2015 22 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Everspring as String
Sat Jan 24 16:23:57 2015 22 {unknown fixed value}
Sat Jan 24 16:23:57 2015 Everspring as dummy Watt meter
Sat Jan 24 16:23:57 2015 0;0.0 {initial value __ not overwritten}
Sat Jan 24 16:23:57 2015 Everspring as dummy counter meter
Sat Jan 24 16:23:57 2015 0 {initial value __ not overwritten}

Check Heater OR Warm Water Supply
For the Nefit Ecomline on point 9 and 11 there is 24V AC when the Nefit is in operation for warm water supply. To Zwave this signal, the switching 24V is feeded to a bridge rectifier and smoothed by a 470 mircoFarad capacitor, then via a 24V-48V DC relay feeded to the input of the Fibaro FGBS001 to bring the status to Domoticz. During no operation and during operation for floor heating this voltage is 0.

Conclusion.
Based on stable version V2.2025 (because it's a production environment).
Blockly is not mature enough yet, to handle Current Power Consumption as comparison value.
Luckily this was not a show stopper because the Accumulated Power Consumption of the Everspring AN158 works as comparison value (while it does not for Fibaro).
Disadvantages of this solution are:
- Why this is working is not understood
- Polling of Everspring must be on
- Everspring needs shortest possible meter report period (1sec)
- Latency of around 12 sec compared to Current Power Consumption
Disadvantage of the Everspring AN158 (compared to Fibaro)
- can not be locked to ON
- unknown ON/OFF state after power down

Todo list:
- deltaT controlled speed (speed 3 or 2): comparison between supply and return; Currently not possible with Blockly
- controlled speed (speed 1 or 0) for remaining heat (in Dutch: restwarmte) optimization
- optimization of Warm Water Supply: first measurements show, averaged per day, that there are 22 switches (!) between Warm Water Supply and Heater
- check the accuracy of both Everspring and Fibaro for curiosity purpose

Current limitations:
- Blockly can not handle comparison based on Current Power Consumption
- Blockly can not parse concatenated strings into variables
- Blockly can not handle comparison based on subtraction of 2 Temperatures
- For Switches there is no Chart Context Menu to download the switch state history in a CVS file for further analysis.

Praise for Blockly
I like to emphasis the importance of Blockly.
For most of us, part of the fun is the making of a domotica system, but for the people around us, it is the use of the domotica system.
An underestimated aspect of Domotica is the maintenance of the system, like slight parameter changes and obsolete hardware replacement.

With Blocky you can develop solutions that can be maintained by others.

Wow, it's a shame that no one else did yet reply to this topic! I really like these projects where self-built electronics is combined with domotica/programming stuff!

I don't understand the workings of the electronic boxes fully. If i understand correctly: When the central heating turns on for the first time a day, you want to run the pump at fullspeed, to deliver a lot of warm water to the floor. Later, when the heater is modulating, you want to run the pump at a lower speed, so the water flows slower through the tubes, and it has more time to give off it's heat to the floor.
This is combined with some Z-wave stuff to tell Domoticz what is happening and to base events on.
Couldn't you do something with a relay that monitors the 3-way valve of your heater (if it has one) so you know when the heater is heating your house? And combine that with some timeswitch, so the pump runs only when the heater is on, and at full speed when the time is between X and Y ?

When i read about the 80 degrees water setting, and the 3000m3 gas usage, i already got a bit irritated. I really HATE it when engineers/technician deliver a installation like that, to not-knowing customers (who need to pay the bills). I have also tuned my central heating system (waterzijdig inregelen). Also did it at my parents house, they have a modern high-efficiency heater (HR-ketel) but the settings on it were horrible! Also 80 degrees, radiatorvalves wide open... Radiator near the heater was bloody hot, and one far away in the livingroom mildly warm. I tweaked all of the valves so every radiator gets just enough hot water. Also brought the heater temperature down to 65 degrees. My parents will save 20% gas, worth about 300 euros! The adjusting of the valves was done at a saturday, it took my father and i about 4 hours to get everything working nicely.
I already placed a temperature activated pumpswitch earlier, the 65W pump had been running constantly for 15 years, what a waste of electricity!

I am quite active in this (Dutch) topic: https://gathering.tweakers.net/forum/li ... es/1619229 Come by and take a look, if you don't know that topic already. Very useful information can be found there, about making your central heating system more efficient. I also like saving on electricity, where a (also Dutch) topic also for exists: https://gathering.tweakers.net/forum/li ... es/1491207
For both the gas and electricity you are sometimes flabbergasted to see how much a simple modification can save you on your energy bills.

Some projects i did:
http://thinkpad.tweakblogs.net/blog/110 ... highcharts
http://thinkpad.tweakblogs.net/blog/106 ... l-database
http://thinkpad.tweakblogs.net/blog/984 ... heeft.html
http://domoticz.com/forum/viewtopic.php?f=4&t=4447

The first two projects aren't used anymore, as they are replaced by Domoticz (i didn't know Domoticz back then). But i learned a lot by doing these projects. I am now busy with a ESP8266 module, it is a serial-to-wifi module, with the size of a post stamp. Really awesome, it has a lot of power for such a small module. I already managed to combine it with an Arduino, to create a DS18B20 temperature sensor that sends it values to Domoticz over WiFi See the topic: http://domoticz.com/forum/viewtopic.php ... 358#p32257

Don't hesitate to share these kind of build-logs, i really enjoy reading these kind of topics!

Apart from the positive story, some comments, don't hate me now :

1. In this picture you use a brown wire for L(ive) buth also for N(eutral). Don't do that! It is confusing for someone who has to work at your creation (or if you need to change something yourself after a few years, and can't remember how something was configured).
2. And please resize the images before uploading. It is now annoying to constantly scroll vertical&horizontal to view the picture

Oh and i think a thermal camera could be a nice addition for your tools collection. I recently discovered a cheap thermal camera, the 'Seek Thermal'. It was available for $199 dollars (they raised the price to $249 now ), and can be connected to your smartphone. With that, you can literally see the floorheating tubes running throughout your house. I am planning to import that camera from eBay in a while. Imported from US to the Netherlands will bring the total price to about 210 - 250 euro's. Not expensive for such a great tool! A professional thermal camera is the same price as a small secondhand car

Hello Thinkpad,
Yes, I have followed your enthusiastic contribution in the different fora and your blog.

Yes I have made donations to the incredible Domoticz team (part of my savings is for them)
b.t.w. a good initiative to promote this.

Yes, you are right with the Blue on Brown connection: that was how it was installed in the house. It's the connection to the safety temperature switch (over Temp.) of the floor.
A cable with 1 Brown and one Black was not in the inventory of the installer?

Yes, central heating tuning is a big cost saver. It all starts with good measurement. I use 16 DS18B20 collected by 4 FGBS001 to measure with Zwave the most important spots.

Yes, I have used a pump switch. But in my house with a 0.1C/hour decline only (in 7 hours from midnight to breakfast the temperature goes down with only 0.7 degrees) the hysteresis of a pump switch is way too large. All I do is maintain in the morning and evening the energy in the floor.

Yes, you are right with the picture size. As you can see in the picture titles I already tried to reduce them, but they need the size of a big stamp

Have not checked your thermal camera thing. I could be a trick in the near IR (just above the Red spectrum) instead of a real IR camera.
Edit: Not a trick: it's Long Wave Infrared 7.2 – 13 Microns

To the subject:
Time is not a good differentiator for the pump speed control. As you can see in the 2 pics above (Stepwise burning versus boost)
you can see that the interaction between the modulation of the Moduline 400 and Nefit Ecomline is complicated and hard to predict.
A combination of requested roomDeltaT, previous experience of the Moduline 400 (self learning), and heater DeltaT.
Typical values for my configuration measured at the Nefit (supply versus return): Tout 35C = Tin 25C up to Tout 45C and Tin 35C.

The pump is changed from a mechanical 3 speed, to a Z-wave 3 speed.
The box has 2 Z-wave switches that controls 2 relays that creates the logic to control the pump (0,1,2,3). Plus one switch to override all (to keep it WAFfed )

My objective is to start fast (heat distribution in the floor) then slow down (to optimize DeltaT). Then slow down more, to pick-up the remaining heat and stop as soon as the floor is hotter then the remaining installation. While during summertime: prevent that remaining heat from hot water production goes into the floor.

Thanks for your response
Domosapiens

Change log:
Added: Typical values for my configuration measured at the Nefit (supply versus return): Tout 35C = Tin 25C up to Tout 45C and Tin 35C.
Edit: Not a trick: it's Long Wave Infrared 7.2 – 13 Microns

Some by-catch (bijvangst in Dutch)
or how to understand hot water usage with Domoticz.

Thanks to Domoticz I'm able to measure the utilization of the valve for hot water production.
On the Nefit there is the possibility to control a valve for an external boiler. This 24V AC contact is rectified, smoothed and given to a relay and that gives the Fibaro FGBS001 a potential free contact: heating or hot water.
I did two measurements of 5 days (with some overlap so statistically not fully correct!), with an outcome that surprised me: 16-17 times per day the Nefit is producing hot water.
1x in the night to keep the internal 25L boiler on temperature ... but at least 15x for our 2 persons household! The total duration of water production was 7.3hr during the 5 days or 1.4hr/day.

The first graph is the frequency distribution of the duration of water production during 5 days (95 samples). There is a rather clear distinct between kitchen, bathroom and shower usage.
(I have to learn to shave myself faster, and shorten shower time. My wife does better).
1/3 of the number of water requests comes from the kitchen and taking 1/6 of the water production.
The second graph shows that 35 times within 5 minutes the Nefit switches on for a next round of heating.
On the cost side:
From my summer usage I know that we use 20m3 gas per month.
With colder water and a larger deltaT this would be some more outside summer season.
But let's say 12x20=240m3 for hot water, around €180 plus electrical consumption by the Nefit for hot water is 41kWh/year, around €9
1/6 of those cost are allocated to the kitchen..... but
Because of 15m distance, there is a hot fill boiler in the kitchen, that takes 280kWh/year, or €62. That brings the total cost for hot water production for the kitchen on €93.
This information helps me to think about a future configuration for floor heating and hot water. My first conclusion, a large boiler and and boiling-water tap. My current cosumption is more than a quooker will do. Hope this inspires you to measure your set-up.
Regards
Domosapiens

It's a long read, but well written.

We also had the standard 80 degree setting and such, until we had a different person for the yearly maintenance.
He said we were wasting money and comfort and lowered the temperature considerably.
Never checked the actual savings but with an yearly usage of around 700m3 it would be hard to notice any real difference.

Have you checked the potential savings in exchanging the pump? You converted an old and already inefficient pump.

Have you checked the potential savings in exchanging the pump? You converted an old and already inefficient pump.

Yes, I did.
Current winter day pump usage has a duty cycle per day of 27% and is 0.40 kWh/day.
That is a mix between off and speed 1, 2 and 3.
A new Grunfos ALPHA 2L would do around 0.16 kWh/day
With this winter day duty cycle it would take 10 years for a break even (pump price is €175)
If we take summer duty cycle into account this would be even much worse.

And it would mean a third Captain on the ship.
The Nefit thinks that it's intelligent, as the Thermostat thinks and then an intelligent ALPHA 2 adding to it ?
Sorry, I prefer to be in control with Domoticz, and that works great with this old pump modification.

Great reflection!

Saving energy is not always done by buying the newest stuff. If your pump was running 24/7 it would have been a smart choice, but in your setup the pump-power scheme is already optimized.
Better spend the money that an Alpha2 would cost on some other nice home automation stuff

Very important to keep realistic, to prevent wasting money

A post on tweakers got my attention
Nefit comparing a new heater with older ones:
http://www.nefit.nl/professioneel/over_ ... nvoorbeeld

Thanks to Domoticz I know some figures of the 28.8kW Nefit Ecomline:
Age: 19.5 years ! Some 5 more years technical life are still there.
In standby-modus: With the current weather there is 1 large morning heating cycle per day of 4-5 hr on 28C to max 35C.
If the sun did not do his job, at the end of the day some additional short 28C cycles are given to the floor.
Plus you see a number of short start-stop cycles for hot water.
Electrical usage during operation is around 87 Watt. Whole day electrical usage is around 0.9kW during winter operations (no figures yet for summer operations) I hope that the energy glitch (now&than) will be solved when I move this weekend from version V2.2025 to V2.2284.
(a semaphore problem during the hourly backup?)

Thanks to the remote pump control (see start of post) and Domoticz, the ground floor pump is running on 2 or 1 or stopped.
Speed 3 was only used in times that it was below 0C outside.
The second floor heating group is running between 1 or stopped.

Gas usage with 4C outside is around 10m3. Overall 0,50 m3/graaddag in a 250m2 and 550m3 house build in 1995.
During the night, (=heating off) temperature goes down with only 0.1C/hour !!! But still I need 10m3 gas to keep it that way !!
I need to find out how those 10x9kW=90kW are lost.

Duty cycle of the heater is around 6/24 =25%.
During operation the heater is modulated back to around 50% of it's capacity.
With the current weather a 8x overcapacity!

Thanks to Domoticz I have now the information to re-engineer for a new solution.
The major is calculate how the 90kW loss can be explained and how to reduce,
and then:
Thinking of a cheap heater with half of the current capacity plus, 2x a 200 liter boiler (one comfort water, one floor heating) with heat pipes as feeder.
After pre-heating the comfort boiler, the remaining sun energy goes to the floor heating boiler.
The hot comfort water need, will not dominate the capacity of the heating installation anymore!

Some next episodes:
- how to measure on 16 points persistent temperature with Domoticz
- how to save 700m3 gas by measurement and tuning with Domoticz
- stay tuned .. for more

LOL, they mention 30W for a central heater in standby. That is quite crazy

Thanks Domosapiens for these great posts ( & Thinkpad & others for their contributions). Saving energy is about gaining knowledge in the first place, these posts pose a great contribution

@Thinkpad, you write "When i read about the 80 degrees water setting, and the 3000m3 gas usage, i already got a bit irritated.".

With an (almost) all-floorheating scheme as Domosapiens house and mine, it indeed seems overkil. But i dare you to reflect on how inefficient this really is. Try to follow my way of thinking:

it seems ridicoulous to let the boiler burn at 80 degrees watertemperature when all you do is throttle it back to 40-45 degrees before being fed to the floorheating. But that water gets distributed in the pipe system to any radiators elsewhere in the house also. If those radiators are shunted through thermostat valves, what you end up with in this scheme is that non-floorheating parts of the house are warmed up quickly, until the radiator valves kick in and 80degrees waterflow is reduced. The boiler will keep producing 80degree water, but it will take less gas because a) waterflow is reduced so b) the return temperature will rise quicker, making the boiler reduce power. So, when the non floorheating rooms are warmed up, all negative effect that remains in producing 80degree water is loss of heat in the pipesystem to the rooms. But also that heat is still dispersed in the house itself (as long as the pipes do not run under the floor). In that respect, it becomes less ridiculous to heat at 80degrees, it could even be a deliberate choice when you want to warm up non-floorheating rooms more quickly for relative short periods of time?

To sum up, is negative effect of heating up with 80 degree water in a floor-heated house not limited to radiaton loss of pipework only?

After more than a year of experience, time to share some info.

Domoticz (even V2.2365) is running rock-solid on Windows.

OpenZwave is running good, as long as you take the propagation rules into account.
(see f.i. http://www.domoticz.com/forum/viewtopic.php?f=24&t=6000
and http://www.domoticz.com/forum/viewtopic ... =20#p38940 )
A concrete obstacle is able to absorb 90% of the signal.
Remember that ZWave has an US origin, where houses are build often different than in Europe !

Currently I'm in a phase of restructuring the temporary (in dutch houtje touwtje) setup.
The GroundFloor setup is now mounted on DIN-rail. The 2 blue boxes contain the FGBS001 4x temperature DS18B20 measurement.
The antenna can be seen under the box.
Signal has to pass trough the wall plus 10 more meters.
On the left theZWave 4-speed pump selection (speed 0,1,2,3) with Philio PAN04
The FirstFloor still needs the restructuring but is a good picture to show how NOT to install: The optimal signal line-of-sight is trough the wall left under an angle of 45 degrees, then 1 meter free space,
then under an angle of 45 degrees through a concrete floor with at least 2 concrete steel mats (with around 15 cm open space while the ZWave wave length is 17cm),
then 12 meter open space.
Alternative ways in the ZWave mesh are:
- straight going down to the first floor (trough the aforementioned concrete) and follow the path of the ground floor system
- go right under 45 degrees trough a window and a wooden roof structure toward an other ZWave hub on the ground floor and then trough 14m free space and one wall.

On the right signal is blocked by the washer
On the back the signal is blocked by the black isolated buffer vat
On top there is the heater.
On the left a number of metal pipes in front of the wall into the line-of-sight direction

.... and if my wife stores the vacuum cleaner in the front ...
ZWave starts with multi transmission errors that break trough the error-check and are interpreted as wrong values f.i. CO2 value instead of a temp value.
These errors can be found in the log and as new lines for that device in the zwfg_0xnnnn.xml.

The FirstFloor ZWave units will be reallocated and mounted on DIN rail also as soon as I can find some time. (will picture that)
Also I will streamline the Blockly control structure now that I have an all season experience in measurement(!) and control.
Will post later about that.

Hope you enjoy,
Domosapiens

Great project you've got here, I've just found this thread, and this is just to let you know that I am impressed.

Thanks CoffeeEric, (and all other 1799 viewers)
Now that wintertime is coming again, it's time for some new experiments.

I'm planning an impulse responsiveness (step-response) test between input and output of the floor heating system:
sending a peak of hot water trough the system and measure the delay (the _t or tau of the PID control) before it comes to the output.
I will measure that under tree conditions:
- with an established water flow in the floor
- with the pump switched from 1 to 3 at the moment when the impulse is feeded to the system
- with the pump switched from 0 to 3 at the moment when the impulse is feeded to the system
In this way I like to understand how much time it takes before the flow trough the whole floor is re-established.
I have an assumption that switching off the pump between the burn-cycles of the heater is not optimal.
It could be that a sustained low-flow (pump on 1) combined with pump on 3 during burn-cycles is a better solution.

Since the latest Domoticz release a comparison on temperatures is possible.
That brings me to my next experiment: a combination of the wall thermostat with the floor-out temperature as control for my heater.
The idea is something like heater ON when water-out is lower then 22C and heater OFF when water-out is higher then 23 (the values are depending on the step-response as measured before).
The wall thermostat will override this process when there is enough warmth from the sun trough the windows.

The objective?
Get understanding how a sun-boiler or A/W-pump could fit in my next heating solution.
My current Nefit heater is 25 years old, but still going strong (thanks to the low temperature heating concept?).

Thanks to Domoticz I have more understanding in the last 2 years, than in the 23 years before!
Will keep you informed,
Domosapiens

Domosapiens wrote:The objective?
Get understanding how a sun-boiler or A/W-pump could fit in my next heating solution.
My current Nefit heater is 25 years old, but still going strong (thanks to the low temperature heating concept?).

Probably not because of the low temperature but because it is a Nefit.
My parents have a 1996 Nefit which is also still in good working condition with minimal maintenance.

Last year I measured our total gas usage in the summer. We use induction for cooking so gas usage was only for hot water.
I extrapolated the hot water gas usage to a whole year being just something like 250m3.
Which in no way justifies the 3000+ Euro expense for a solar boiler setup.

That said, relatives of mine have a solar boiler and it does work pretty well.
In the summer months they hardly use any gas.

Thanks for your heads-up bbqkees.
You are right about the hot water gas usage, that does not justify the investment.

My objective is to use the sun-boiler in the hot water AND floor heating chain with 2 separate boilers.
For my floor heating around 25 degrees sustained heating is already enough.
It will be a tilt-able construction on South: flat in the morning, tilted during midday.
Low tilt during summer, and around 60 degrees during wintertime.
My current estimation of savings is around 450€ per year (Eastern part of the Netherlands).
All controlled ... by Domoticz!

My 2016 project ..
Domosapiens

Time for some updates:
The first floor control is organized now. Left top unit the remote control for the FirstFloor pump and the Thermostat_bypass .
Both with 1x Philio PAN04-1.
Control comes from Domoticz.
With push button for local test and a WAF switch to override all remote control

Left second unit: Nefit (=central heater, in Dutch CV) pump_running detection with LED: senses pump activity converted to a digital signal for FIBARO System FGBS001.

The Blue boxes contain the FIBARO System FGBS001 with 4 temperature sensors and 2 digital inputs per unit.

Digital inputs are used for:
- Nefit pump running?
- 3Way valve in Hot Water position?
Status is used by Domoticz

Temperature inputs are used for:
- CV_IN, CV_OUT, CV_Cold-Water_in, CV_Hot-Water_out
- Floor-IN, Floor-mixed, Floor-Out
- A few specific groups like bathroom
All temperatures are handled by Domoticz

Status monitoring of the Nefit via IPcamera: The Nefit has 2 options for connecting a thermostat:
- The upper 2 connectors are for an ON/OFF thermostat.
- The second 2 connectors are for the Nefit specific modulation, in my case Moduline400. Domoticz controls the ON/OFF thermostat as Thermostat_bypass.
Why? We will see in the next post.