Overthinking the controls

[PeterW]

Right.. so following the post where I put the planned controls and layouts I have been thinking about the ASHP temperature control mechanism for this.

 

You can see from the schematic there are two pipe stats that are directly linked to the ASHP control box. In the ASHP instructions, this is used to regulate the water temperature in the system, and as I want 2 different temperatures then I need 2 stats. The unit is supplied with a standard Siemens RAM1 pipe stat which seemed ok until I checked the differential - 8K +/-3K...!!! So at 37c set point, I could get a range of 32-42c which seems a bit excessive.

 

I am trying to keep this as "off the shelf" as possible, and looking at specs its clear that 4K differential is the norm on these things unless you go upwards of £35 per device. I can get down to about 2.5K with one of these Computherm Capilliary Stat which seem pretty good. ( @JSHarris didn't you use Computherm..?) but I'm still querying if it is "good enough".

 

The house isn't passive by any stretch but will have good air-tightness and a well insulated roof, with a January heat load of around 1180KwH. Overshooting the 35c stat will only result in the buffer being warmer than needed as the UFH will be set to 32c flow.

 

I could go with something such as one of the many £8 DTC-100 based controllers from China and get myself 0.5C differential control on the water temperature but I do wonder how the ASHP will cope with this, and what the benefit will be as there is 9m of pipe between the ASHP and the sensors.

 

I'm not going down the custom code / build route on this but I'm trying to build the best "analog" system that I can without spending a fortune but I'm left questioning how good do I need to go..??

[TerryE]

1 hour ago, PeterW said:

I'm not going down the custom code / build route

 

Chicken ;-?

[Ferdinand]

I have not reached a view yet, but i do wonder about the impact of complex control systems on the next owners of a house.

 

I have seen some on BH that would scare me off from buying as my perspective is that really I just want something I can leave alone except for twice a year, and has no moving parts if possible - while delivering perhaps 80-90% of the performance that would be achievable by adjusting very often.

 

I think that a degree of complexity delivers, but that then needs a maintenance and repair strategy that is sufficiently simple that will last the lifetime of the system. Alternatively there could, I suppose, be an easy "strip drown and simplify" strategy for when it comes to be sold.

 

For me, buying custom controllers from China, whether with custom code or not, may be a step too far.

 

I appreciate that many on BH build with a "house for life" (*) expectation so that does not apply.

 

Ferdinand

 

* (Favourite anecdote.) Or, as Lord Morris of Castle Morris put it in the early 1990s after he bought Foolow Manor in Derbyshire in preparation for retirement: "I have bought a small manor house in Derbyshire to decline and die in" - and he did. Matthew Parris is now doing something similar.

Edited April 1, 2017 by Ferdinand

[PeterW]

7 hours ago, TerryE said:

 

Chicken ;-?

 

Yep...! 

 

Its more about supportability and maintainable components when it's not just me in a house. I've got a number of ducts in for DS18x sensors to monitor what the house is up to but I need to make sure that some of the components are easily replaced. 

 

It annoys me that some of standard components on items such as UVC Aquastats are of similar spec too as it means if I want to tighten up control of all of the tolerances I need to discard supplied components and move to more custom elements. I noticed ESI have brought out a digital aquastat now but the majority of UVC controls are still bimetallic stats that haven't changed in 40 years. 

 

Oh and it doesn't help when the other occupant likes to use the "off" switch on controls rather than letting the controls do thier job .... I'm thinking of putting it all in a locked Perspex box with just a small hole for a finger to press a button if extra hot water is needed ....

 

[Triassic]

I was involved in the installation of a new heating system in an accommodation block at a Scout Camp Site. We found that customers were fiddling with the controls so much that we moved them into a locked cupboard and replaced them with a dummy room thermostat, this they could twiddle to their hearts content. 

 

I'm a fan of KISS, at the end of the day you have two points of human interface, "is it too hot/cold in here", or "is this water too hot/cold"? In both cases a bit of tweaking of the controls will bring these into control.

 

 

Edited April 1, 2017 by Triassic

[Jeremy Harris]

Yes, I used Computherm room stats, one for heating, one for cooling.  The ones I used had a 1 deg C hysteresis as standard, that was adjustable to 0.1 deg C by moving a link on the back.  They were these ones: https://thermostats4u.co.uk/products/computherm-q3rf

 

For the buffer tank stat I used a standard mechanical remote probe thermostat, a Drayton I think, that has around 5 deg C hysteresis.  That seems to work well on the buffer, as it cuts out at 40 deg C and back in at about 35 deg C.

 

[PeterW]

24 minutes ago, Triassic said:

I'm a fan of KISS, at the end of the day you have two points of human interface, "is it too hot/cold in here", or "is this water too hot/cold"? In both cases a bit of tweaking of the controls will bring these into control.

 

 

I entirely agree hence my post ..! My issue is the delta range on some of these components as they are huge when you are considering such low temperatures. 

 

The idea with running the floor at 32c and the buffer at 35c is that in the event the buffer becomes exhausted the ASHP will be feeding pretty much directly into the UFH (albeit with the buffer acting as a low loss header in this instance) and this is only really likely to happen when the temperature  is very cold. The concern I have though is that I'm at the mercy of a number of components and their tolerances :

 

Room Stat +/- 0.5K

Buffer Stat +/- 5K

Pipe Stat +/- 5K

Blend valve +/- 1K

 

Its fine in this instance if the buffer and ASHP are above the temp required for blending the floor but not if below as the floor will never reach temperature. But taking the ASHP up to 40c from 35c starts to erode the CoP and impact frost cycles, and then the same is true at UVC temp and I could be up above 52c with a CoP through the floor....

 

Temptation is to put 4 of the DTC-100 devices in a stack and use them to do all temperature control.   At least I can set the tolerance to say +\-1K and they are cheap enough to be disposable.

 

The downside is that I get 4 glowing displays and a temptation to fiddle ......

[Triassic]

I agree, the delta range is wide, but I thing you will find that once you've got the system up and running and with a bit of fine tuning it'll be a set and forget system. The only time I fiddle with our system is as the first cold days of winter sets in and the incoming cold water temperature drops.

 

 

Edited April 1, 2017 by Triassic

[Jeremy Harris]

The reason I decided that fine control was necessary has a great deal to do with the thermal response time of the whole house.  I've found that putting even a couple of degrees more heat in, as a result of a high hysteresis room thermostat, means that the house room temperature overshoots, just from residual heat in the slab continuing to heat the house for several hours after the heating has turned off.  Being able to shut the heating system off at precisely 20.5 deg C, means the room temperature only overshoots by less than 1 deg C, so stays below about 22 deg C during the overshoot period.  Even in the coldest weather our heating only fires up once in the morning, and never comes on after mid-morning, as the residual heat in the slab just keeps heating the house until the next morning, a bit like a storage heater.

 

The Computherm Q3RF room stat is about as simple as these things can get, with just a temperature setting and a small slide switch to select normal or set back temperature (both temperatures can be adjusted) . 

[TerryE]

My general issue is that I have to control of  house where the winter losses are of the order of 1kW or less and the number of net heating days maybe 90 or so and the number of net cooling days is possibly greater.  Doing this is actually going to be fairly easy, but the reality is that this is such a niche market that there aren't any off-the-shelf control systems to do this.

 

Jan was asking me why I am proposing to use so many thermometers; did I need them all?  My response is that I think that I actually only need three for the control, but that the rest are logging only to validate that the control is working effectively and in some sense near optimum. By going custom, I can quickly and cheaply build a system that does what I want and I can also tweak it post go live in software rather than having to reconfigure hardware and buy new bits and pieces.

 

Once the system is stable and working, I've then got a few years to let the off-the-shelf market catch up and I will know exactly what I am looking for.  

 

In the meantime, so long as I have a complete stock of LRUs (line replaceable units. -- pluggable spares) and my son-in-law understands the system, then there isn't a maintenance risk for Jan.

[PeterW]

1 hour ago, TerryE said:

Jan was asking me why I am proposing to use so many thermometers; did I need them all?  My response is that I think that I actually only need three for the control, but that the rest are logging only to validate that the control is working effectively and in some sense near optimum.

 

So this is my thought too @TerryE - my issue though its the 4 control thermometers have such wide range of hysteresis !

[PeterW]

2 hours ago, Triassic said:

I agree, the delta range is wide, but I thing you will find that once you've got the system up and running and with a bit of fine tuning it'll be a set and forget system. The only time I fiddle with our system is as the first cold days of winter sets in and the incoming cold water temperature drops.

 

 

 

I think thats my plan too - and I will set the stats slightly high and see where they end up after tweaking. I've got an accurate digital thermometer anyway that I will use to check the calibration before I start.

[Jeremy Harris]

FWIW, I have a couple of the DTC-100 units and they have 0.1 deg C hysteresis, but 0.5 deg C accuracy.  They would make pretty good pipe or tank stats, I think, as you can adjust both the on and off set points.

[TerryE]

I am getting better than a repeatable 0.03°C accuracy out of my calibrated ds18B20s.  It's just a lot easier when you are working to a known accuracy, IMO. 

[Jeremy Harris]

5 minutes ago, TerryE said:

I am getting better than a repeatable 0.03°C accuracy out of my calibrated ds18B20s.  It's just a lot easier when you are working to a known accuracy, IMO. 

 

It is, and my thinking was similar to yours, instrument the house and collate a fair bit of data, then look at how to best control things.  The only problem I had was trying not to get sucked down into the sort of "continuing complexity vortex", just because it seemed like a fun thing to do. 

 

The system I've ended up with wasn't fun or challenging to put together, but does have the advantage of only using three, off-the-shelf thermostats, plus four plug-in DIN rail relays, all of which can be replaced pretty easily.  Despite the simplicity of just using a room temperature thermostat to control the whole house temperature, in practice it does seem to do a very good job.

[TerryE]

Jeremy what I didn't want to get into is trying to out and find OTS (off-the-shelf) control components, put them in and then find that I needed to rejig everything and by more bits.  OK, I am also a software geek which helps. Hence my two phase approach:

1. Gather lots of data.  Have some basic on-off control output, and do the algos in SW until I've fully characterised the system and have confidence that it works as I anticipated.
2. Once I know how my house behaves and how to control it, then do a second phase productionising / procurement of an OTS solution.

To be honest, watching / tracking your learning curve was one of the reasons that I decided to go down this route.  The main advantage of this split, IMO, is that (2) isn't time critical as the house will be running happily.  I can take my time over finding the right kit for my needs.

[Guest Alphonsox]

The Computherm stats are also available from Amazon, just picked up one for the UFH control.

https://www.amazon.co.uk/COMPUTHERM-Q7RF-Wireless-Programmable-Thermostat/dp/B00SI1D6FY/ref=sr_1_1?ie=UTF8&qid=1491050748&sr=8-1&keywords=computherm

 

[PeterW]

24 minutes ago, JSHarris said:

The only problem I had was trying not to get sucked down into the sort of "continuing complexity vortex", just because it seemed like a fun thing to do.

 

The exact reason I'm looking at alternatives ..! 

 

I am thinking that having a single control panel built of off the shelf DIN rail components and the DHT-100 units would be fine, but it would put a few people off I'm sure when it came to selling. The benefit from direct off the shelf components used "as designed" is that if something does fail then it's an easy replacement. For consistency I'm looking at using all computherm components though so I should be able to get some standardisation on a "known" temperature. 

[Jeremy Harris]

I have two Computherm Q3RF 'stats on the wall in the hall, one above the other.  The top one is set up as the cooling 'stat, the lower one as the heating 'stat.  Generally they both are within about 0.1 deg C of each other, and both seem to be very close to the temperature indicated by an NPL calibrated mercury lab thermometer I've got, so I think they are fairly well calibrated as standard. 

 

It's a pity that Computherm don't make a remote sensor version of this 'stat, as that would then mean you could have a fairly common set up for both air and fluid temperature control.

[TerryE]

22 minutes ago, PeterW said:

I am thinking that having a single control panel built of off the shelf DIN rail components and the DHT-100 units would be fine, but it would put a few people off I'm sure when it came to selling.

 

Peter, I've just spent the most of the last three years building my bloody house, so the last thing that I want to do is to plan for selling the bloody thing.  Preparing for that eventuality is very much a quite separate phase for me.

 

22 minutes ago, PeterW said:

The benefit from direct off the shelf components used "as designed" is that if something does fail then it's an easy replacement.

 

In a way, I think that you are 100% wrong here.  If you need to replace any control component these days, then what is a realistic spares and replacement life?  2 years? 5?  If some bit of kit goes wrong in 3 years time then I suspect that you might  have to do (major) work to replace it.   Forward buying is a expensive game if you are only have one of each and they cost a few hundred £.

 

Jeremy and I have both work with the armed forces and understand their logistics models.  They easiest way to guarantee a service life is to buy a sufficient spares holding at the point of commissioning / procurement that you have a set of plug-in replacements for any sub-components likely to fail.  So in my case my basic IoT compute module currently costs  £2.61.  It plugs into a 2 × 8-pin header on a passive board, so is a quick swap out. I think that I can afford 10 spares; just in case since I use the same module in all of my IoT sensors and controllers.  

[PeterW]

Hmmm ..... by control component I mean the ones with moving parts that could wear. 

 

So I have :

Pipe Stat

3ch Controller

Room Stat

Diverter Valve

Zone Valve

 

most of that lot has been "standard" for the past 15-20 years with the exception of the 3ch controller, and the speed that UK Heating Plc changes its technology I doubt they will much different in 5-10 years time. 

 

[Jeremy Harris]

My approach to maintainability was a bit of a mix.  For some things in the house, I have just bought a fairly large stock of spares, perhaps the best example being the UV water disinfection unit, as the lamps are hard to source for a reasonable price now, two years after I fitted the thing.  I've got a ten year stock and have accepted that the whole unit will almost certainly need to be replaced by then, as I doubt the right size lamps will still be available.
 

For the heating/cooling/DHW system I took the view that it needed to be able to be repaired or maintained by any reasonably competent heating engineer or plumber, in the future.  To that end the thermostats could be pretty much any dry contact thermostat, and the relays are a very common standard DIN rail component that should be around for decades.  I've documented things with detailed wiring and layout drawings, plus a bit of text describing how it works.  My main worry with the microcontroller system I spent many, many hours developing, was that it was very unlikely that any heating engineer would ever be able to figure out how to diagnose a fault, let alone fix it.

[SteamyTea]

If you are designing/building a custom control system, is it worth making sure that basic off the shelf controllers can be easily fitted if needs be.

 

So take a room thermostat.  It is fairly easy and cheap to design one that can control a switch with great accuracy (even add in weather compensation) and log data (and even publish that to a website).  If in years to come, someone else needs to change that controller, it would be good that they can go to B&Q, get the most basic mechanical thermostat and just wire it in.

Some marking up of cables would be necessary i.e. power to controller, control lines, but all cabling should be marked up anyway.

 

Maybe something worth thinking about.