Boiler optimisation in solar systems

[dinitro]

Been grappling with this for some time... just wanted to know other peoples thoughts

EW panels
1 E + 2 W
Thermal store approx 300l
Boiler heats whole of thermal store, contents used to circulate heating circuit.
Boiler control was timer + cylinder stat clamped to boiler return pipe.
boiler only fires due to time and if return stat less than 50 degrees.

problem: in summer boiler still momentarily fires

New solution; capillary stat in top pocket (set to 65), capillary stat in botton pocket (set to 40),
cylinder stat on boiler return pipe, timer all wired in series to boiler control (boiler
powered seperately).
This has resulted in the boiler not firing at all and not even causing losses due to activation of pump.
Only when all conditions are met does the boiler fire.

dinitro

[Antman]

Hi dinitro

This is also why I installed two cylinder stats on separate zones.
I now think that two boiler coils (a smaller high level coil and and 'normal' level coil) are also an option for minimising boiler input so that you can put a small quantity of heat into the very top part if needed during those so-called "summer" days when you know there will be enough solar insolation later! Then a second zone valve and lower stat would control the lower coil.

This also ties in with the currently utopian requirement for a decent weather compensation controller that can be configured for user requirements according to system components, location, weather trends etc.

Antman

[trubble]

I still don't have my store here, but have been 'collecting' bits and ideas - to stop the boiler nuisance firing, two stats is perfect - I have read that if the rads or water (via plate X) energy is removed from the store directly, a blast of cold water past a single stat can mean a 5 minute firing.
What I want to do before I get solar, is charge the majority of the store with the boiler, almost to the bottom - so will have a top stat set to, say 50 degC, and a lower stat set to, say 60 degC - making the boiler fire when the store is really depleted, and then fill the whole of the store almost to the bottom. Temps sound the wrong way round, but it means more store gets heated per firing = more condendsing = less gas. Hopefully!

I tested a built-up circuit yesterday on the kitchen table, with the wife shaking her head  at me - and it works for less than a tenner from maplin for the power relay and base - prob same as the DPS twin stat circuits.

Both stats off, boiler off, lower stat calling boiler still off, both stats calling switches boiler on, lower stat satisfied boiler still on, both stats satisfied boiler finally off.

Needs a relay, and due to my hopelessness with electronics I went here - http://www.dutchforce.com/~eforum/index.php?act=ST&f=24&t=19643 and asked nicely   

Swap terminals 1 and 2 on the diagram and it appears to work.

I'd be very interested to hear from DPS owners to see if they have a similar method of using two stats?

[David]

I have the boiler controlled by a TDC2. In summer I have the boiler set never to come on automatically, I can turn it on by hand if necessary. I see no need for anything more complicated.

I have no idea if one of the more expensive TDCs can control one E and 2 W solar panels and a boiler, but I assume so.

[kristen]

"it means more store gets heated per firing = more condendsing = less gas"

Once the store is hot, but if the radiator circuit is still returning cold/cool-ish, can you carry on the boiler condensing cycle by heating the CH circuit?

[wookey]

Is it just me, or is this another example where a few sensors and a smart programmable controller would be a nice solution? Using multiple thermostats, as dinitro has done, works, but you can only change the trigger temps by manually adjusting the stats, and the logic by re-wiring. Much nicer to do it in software: "IF (tank_top > 65 AND tank_bottom > 40 AND boiler return < 50) THEN boiler_on".

This way also avoids having all the temp wiring be mains.

[trubble]

"it means more store gets heated per firing = more condendsing = less gas"

Once the store is hot, but if the radiator circuit is still returning cold/cool-ish, can you carry on the boiler condensing cycle by heating the CH circuit?

Don't forget all this is pie in the sky so far, until I actually have the set up fitted and  working - but from reading around all this stuff, if the rad circuit kicks in (there's close to 20 rads across both of my heating zones) I reckon store temp will drop fairly rapidly, so I'm guessing unless hot water is drawn off, the only time the boiler will fire is shortly after the rads calling for heat. I'm in two minds as to whether it's better to have the rads completely separate from the boiler controls, with their own pumps, valves and stats, or electrically linked in some way so the boiler fires only when rads AND store are calling for heat.
The 'no boiler unless return is <50 degC' is a seriously clever plan too, duly noted and added to my wishlist...

Programmable logic controller idea I think is fantastic - I have no idea as to the way to do it, but the possibilities using a controller with a few temp probes (eg ambient outside temp, return to boiler temp, even the incoming air temp from a hvac unit) would be absolutely awesome... if I could get a grown-up to help me with programming one.
The possibilities are endless, unlike my electronical knowledge.

Can it be done?

[wookey]

Can it be done?

It is done. The electrics part is very simple (1-wire sensors). The actuating part is also pretty simple: 1-wire actuators or I2C digital IOs to relays or mosfets. The computing part isn't too hard although there are millions ways to do it.

The hard part is making a flexible but comprehensible user interface so that normal people, or at least plumbers/DIYers can get a box, plug it in, connect up some sensors and pumps/valves to the outputs, then configure the thing to do what they want.

That 'configure the thing' hides an awful lot of ifs and buts. Imagine you have installed 20 sensors. You need a fairly painless way to work out which one is which and give them useful names. Then you need some generic way of displying the results and setting up the control logic. Something with a picture would help a lot. Being able to do it without standing in the airing cupboard might be nice (exposing a web interface is the classic answer here, but that adds 'networking the airing cupboard' to the list of plumbing tasks, so doing it standalone should be possible).

The closest thing to a decent home heating control system I've found so far is 'DIY zoning': http://diy-zoning.sourceforge.net/index.html which I haven't even managed to build yet (I hate java!), never mind run. It is also currently targeted at US-style blown-air heating, but that can be fixed and it is being used by at least one UK water/solar system owner (Tim Small). It does do a lot of cool stuff, like proper PID control, and may well be part of the answer. Tim is giving a talk about solar heating and free software and embedded Linux on the 6th June in London at OpenTech 2008. I'll be there to wave a solar tube at people to give some grounding in reality.

[dhaslam]

You have to be a bit brave (or foolish) to automate controls for a solar system using your own software.     I am going to compromise a bit and  use the standard controller but monitor performance  using a program. One thing I need to be able to do anyway is to monitor the difference in temperature in two cylinders and only transfer heat when the DHW cylinder is cooler than the other.  Writing your own can take account of things like transferring in early morning but not later when the sun is up.   One rule I have about software is that you allow 10% of time for writing software and 90% to make it work.  One of my (accounting) programs  has 40,000  lines of code and there is probably quite a few places where the answer is arrived at by cheating.