Boiler control questions

[Steve S]

For reasons best known to myself, I'm in the process of developing (for personal use only, not commercial) a heating controller. This will be PIC-based, web-enabled and incorporate features such as weather compensation and holiday mode (for both CH and HW). Yes, I know there are commercial items available, but I get a lot of enjoyment from doing stuff like this.

So far the RTC is running, as are the Dallas 1-wire thermometers for room stat, cylinder stat and outside temp, and web server.

Anyway, I have an idea I'd like to pass by the experts (Drivel need not apply)....

Application is a Y-plan system. As it stands, during the winter when CH is required, boiler runs for CH rarely seem to coincide with boiler runs for HW. Boiler (situated in garage) has to heat up from cold each time. Naturally both CH and HW controls will incorporate (configurable) hysteresis and impose minimum boiler run times. The idea I have is this: Suppose we have CH demand, but HW temp is above HW minimum. When CH demand is satisfied, then I would 'top up' the HW to HW max, thus possibly eliminating an unnecessary boiler start. Vice-versa, of course, for HW demand and CH top-up.

Does this idea make any kind of sense, or am I totally out of my tree?

Steve S

[sponix]

Yes, it makes sense. Whether it'd save much fuel is debatable (imho)

I wonder if it's possible (Using a system of valves) to divert hot water from the HW cylinder to the radiators instead of firing up the boiler?

Surely it's more efficient to 'lose' heat from the cylinder to heating the house than have it lost to the airing cuboard, iyswim?

sponix

[Ian Stirling]

But then you've got the wasted heat in the boiler popes to the coil, when time comes to reheat, and the boiler goes off with warm pipes, and the hot pipes from the cylinder to the heating. It may be in many cases that these are outside the habitable space, and so the heat is wasted.

If you have the room, 3-4400mm of roof insulation round the CH tank will greatly reduce loss.

I expect heat exchanger ventilation would make a fair bit of difference too, assuming that the insulation elsewhere is adequate. Seal all leaks, and actually design the ventialtion rather than letting it happen.

[Set Square]

If your boiler can cope with heating the CH and HW at the same time (if it can't, maybe you should have a W-plan rather than Y-Plan) *and* if the HW cylinder is well lagged, wouldn't it make more sense to - in effect - have HW on constant so that, whenever CH is on, you also heat the HW to the max setting?

On a more general note, it is really worth putting a fancy control control system on a Y-Plan system? Wouldn't it be better to try to convert it to an S-Plan++++, with a separate zone for each room? It would presumably be easy enough to have a temperature sensor in each room. You would then need to control the flow to each radiator - possibly using remote control radiator valves, some of which have micro-switches to feed back their status. You can then have a different heating profile for each room to suit your lifestyle - but still have global over-rides for holiday/party settings etc.

[Set Square]

It would have to be a bl**dy big room!

[John Rumm]

Yup, I see what you are saying.... much of the answer may depend on specifics of your boiler however. A modern modulating condensor will have a different closed loop response than for example a large cast iron lump with all or nothing control.

In the latter case there is probably good argument to save an additional heat exchanger cool down/reheat cycle if this can be achieved with tacking on a HW reheat at the point where the room stat is satisfied.

[Steve S]

S-Plan++++ would be great, but would be very expensive to implement. Floors to be lifted, cables to be routed, redecoration, etc. etc., not to mention the remote control valves. My solution is essentially free except for my time in hacking a bit of C. Boiler copes with CH and HW together except in the coldest weather.

Steve S

[sPoNiX]

True..

I'm thinking of the situation where no hot water is going to be used but the water in the cylinder is hot.

The water in the cylinder could be pumped round the rads, thus ensuring the heat isn't wasted.

sponix

[wildcard]

You're doing all this (inc a webserver) on a humble PIC?

Have you considered a Rabbit core instead? Loads of i/o and ethernet built in....

[Steve S]

It's a PIC18F6621 running at 40MHz with the free Microchip tcpip stack. Already most of the way there with only 40% of the rom, 47% of the ram used, and oodles of space in the external eeprom for web pages/config info. Webserver is already running. Will also add smtp for outage/fault notification later. Rabbit core was *much* more expensive.

Steve S

[Steve S]

"John Rumm" wrote > Yup, I see what you are saying.... much of the answer may depend on

It's the latter currently. Looking to replace probably spring/summer 2007. Updating the firmware accordingly will be no problem :-)

Steve S

[Set Square]

I think the point I was making was that I personally wouldn't be prepared to put a lot of time and effort into re-inventing more or less what could already be achieved with conventional controls. If I were going to program my own system, it would have to do something dramatically better than anything I could buy off the shelf.

[Ian Stirling]

:) Seriously though - once you get over a diameter or so of extra insulation - 400mm on a 400mm tank, the gains from further insulation get quite small. At the wall of the tank, the circumference is 1.2m or so. At the outside of a 400mm tank with 400mm of insulation, it's around 5m, so the effect of an extra mm of insulation is down to 1/4 of that next to the tank.

[Steve S]

I see where you are coming from, but I'm partly doing this for the enjoyment. There's plenty of scope for incremental enhancements later on, without hardware mods to the controller.

Steve S

[Ed Sirett]

I think this is nice project. You can also add features later like:

Fire boiler for 30 seconds each day it's never used to stop the pump jamming while you're away.

[Steve S]

Thanks, Ed. Daily boiler run is already planned, in addition to occasional exercising of the 3-port valve when CH not in use. Also thinking about using NTP (or possibly something simpler) for time updates -- needn't worry about the clocks going forward/back then.

Cheers,

Steve S

[Ed Sirett]

Actually you don't have to anyway as the heating season more or less coincides with the GMT season.

[Steve S]

Agreed, from a control point of view it's not necessary, but from a user interface perspective it is desirable not to have to remember that the controller is always on UTC.

Steve S