Boiler cycling

[Mike]

The recent very cold weather has thrown up a problem I hadn't noticed until now.

Boiler is a Grant 70/90 Euroflame and it's 25 kW output should be more than enough to heat the house through about 20kW rated radiators.

And indeed it did when the weather wasn't so cold outside. But now it's colder it can't keep up. Have traced the problem to the boiler output temperature rising up to the cut-off limit (just under 90 degC) then turning off until the pump circulates the water enough for it to drop to the turn-on value (appears to be around 75 degC).

With the boiler in it's on or off state, the return temperature is about 65 degC plus or minus a degree or two.

Have tried the pump (Grundfos Alpha) in every position on it and all give the same effect, though some vary the time taken that the boiler spends on or off. But typically it is 2-3 mins on, 3-4 mins off. Obviously if I could get it to be on a bit longer and off less the house would be warm again.

Whereas in warmer weather I didn't really notice the radiators warming up and cooling so much now it is obvious that they are continually heating up and cooling down with the boiler.

Arrangement is

|-------->---------------- pump----4 zone valves to HW and 3 CH zones Boiler | | | |

|------------------------ pump----4 zone valves to HW and 3 CH zones Boiler | | | | |--------

[Andy Hall]

Basically the flow rate is too low, as is the return temperature. You have about 25 degrees temperature drop on a design that should be about 12 degrees.

You don't say what pipe size your main 40m runs are, or which model Alpha. Is it the larger one?

One relatively simple thing that you could try is to go and check all the radiator lockshield valves and open them. If they are closed down too low then this will restrict flow. Of course it may be that some have to be quite a bit for balancing, but at least if you could improve the flow it would help.

Beyond that, you could look through the design guide published by the Copper Development Association and do the calculations for the pipe runs.

you plug in the numbers for the main pipe runs, that should give some clue if there is something inadequate.

You can then work back and see if putting in a more powerful pump will help (assuming that radiator tweaks don't)

[Mike]

Partially 28mm, but about 15m of flexible 22mm plastic was needed to poke under a floor we couldn't get under properly. Did wonder at the time if I might need to put in parallel runs of this pipe so there is a rope left to allow more to be dragged through. However couldn't see a way of parallelling these up without causing more resistance. Need a sort of Y joint whereas all I could find were Ts or swept Ts.

Yes

Two are flat out, rest are balanced to maintain all radiators at roughly the same temperature drop.

Thanks - will do. Sounds to be what I need.

[Andy Hall]

You'd have to compare the resistance of adding the fittings vs. having the extra pipe. Also, for the plastic section, you may need to ask the manufacturer because characteristics may well be different to copper.

If the others are closed almost down in an effort to get a reasonable temperature drop across these two, it sould indicate inadequate pipework for the size/distance of them.

You could look up the radiator manufacturer data for the radiators (or ones close to them). There is not an enormous difference between vendors for typical finned radiators. THe main effect is the number of panels and which have fins. If you find something close in size, you can scale it. This should give you a gross heat output assuming the EN testing method. You then need to look at the derating table for the mean water to air temperature correction factor. For 82 degree flow and 70 return, (MWTA about 55 degrees) the normal derating is to multiply by 0.9. So if you determine by size that a radiator has a nominal output of 1kW, its actual output under what should be the design conditions for your system will be 900W. You can then take that figure and plug it into the CDA method and just check that the pipework to those two radiators is adequate for them. A rough rule of thumb is that 15mm pipe should be good to carry 6kW over distances of a few metres (but not 40 :-) ).

You could try opening up a few of the other radiators (note the lockshield positions for setting back) and see if it makes much difference to the flow and return temperature on the main return. If it does then probably the focus should be inside the house. If not, then probably the long pipe runs and you could need a more powerful pump.

It's slightly laborious but you don't have to be hugely accurate.

In effect, the system is adrift by a factor of 2:1, so it should be possible to deduce what is going wrong. You can work back to determine the pump requirement given the resistance and from that see if it is within the range of the pump.

[Mike]

I have been assuming the main reason for the closing down is most radiators are fed from quite short runs from the pump/zone valves under the stairs (shortest is You could look up the radiator manufacturer data for the radiators (or

I think my worst case is the pair of radiators (about 3kW total) at around

20 m there and back of 15mm.

From a first pass I need a head of over 8m and the pump obviously isn't up to that.

Can I put another pump in series on the return to the boiler to see if the problem improves ? (i.e. do pump heads add ?)

Also I appear to be right on the velocity limit of the section of 22mm pipe from the boiler. The article you quoted mentions long term erosion of the pipe but does anything silly happen to the resistance effects of the pipe for a quick test.

I'm thinking the proper solution is a separate pump at the boiler for upstairs and downstairs systems with separate pipework as well, unfortunately still with the 22mm length unless somebody knows where I can get flexible 28mm pipe in 20 m lengths, but I'd like to get as much understood with the current system before changing it.

Many thanks

Mike

[John]

Grant were one of the early entries into the Condensing oil boiler market but are not one of the frequently encountered manufacturers up North at least. Is your model a condenser or a non condenser?

The hysteresis your stat exhibits is perhaps a bit excessive and a suspect stat may be involved here, however your return water temperature appears to show little variation between cut out and cut in on the boiler temperature. This does suggest the rate of flow through the radiator system to be too low.

I think you should get hold of a differential thermometer with sensetive clip on probes or a non contact IR thermometer and check the differential temperature across each radiator under steady-state conditions. If your boiler is non-condensing the differential should be approximately

10/11degreesC.

If you study the FAQ on balancing a system it will help you to understand the aims of balancing a system correctly and if your radiator load is as you mention the performance should be closer to the expectations. Does your assumed load include the hot water demand also or is that on top of your figures? You will find that the boiler will operate cyclically but with a higher flow rate you may find the thermostat responds better because the water moves more "actively" within the boiler, if not consider replacing the boiler thermostat

This can be minimised by matching the boiler output to the load (trivial but needs calibration and analysis equipment) but you will probably find that balancing the system will give better comfort levels and once the house is warmed up you will not be so conscious of variations

[Andy Hall]

OK. I think I'd contact the pipe manufacturer or see if it's on the web site and see if you can get the equivalent figures to the CDA ones for copper. I believe that plastic pipe has slightly narrower internal bore and may also have different resistance behaviour. As a first approximation, you could treat it as copper and plug the numbers into the CDA formulae and see what you get.

It's good that they have their own home runs though.

My gut feel is that that ought to be OK, but I'd run the numbers to check.

I don't believe they do, unfortunately. If you put in a more powerful one then the first is going to end up making that worse I think.

I was afraid it might be. Unfortunately the plastic is a bit worse than the copper as well. THe design point is to stay between 0.5 and

1.5m/sec, although sludging shouldn't be an issue if you are taking care of the system with inhibitor. Beyond 1.5m/sec the noise level increases but that may not be an issue if it's outside or under the floor. However, the head required to increase the flow starts to increase dramatically.

It may be that the 28mm is adequate, especially if copper for this amount of heat. Another option would be to run in two pairs of

22mm from the end of the 28mm and split those upstairs to downstairs. it would be reasonably easy to calculate for that, but you could still be towards the pump limit.

A second pump and pipe run would be the other option, and you could be pretty confident of each working and adjustability for each as well.

I have never seen barrier pipe or even non-barrier in 28mm coils. You could contact Hepworth or John Guest and see if they can do it to order.

[Mike]

Non condensor

temperature.

What should the stat limits be. The upper limit is programmable via a knob on front of boiler but manual doesn't say if the lower limit is fixed or moves with it.

Am using a Maplin non contact IR. As the system has three CH zones I think in better weather it got the system up to temp and then everything was ok. But now under heavy load it can't get there.

That was what I used. Most helpful. I thought I had got the system to around 80 deg in, 65 out for most radiators but as it turns out this was only when the room thermostats were being overridden one by one.

There is provision for a hot water feed but it is currently not activated as I am going to be using a small hydroelectric generator to drive the immersion heater so hopefully the boiler won't be called upon for this.

Thanks

Mike

[Mike]

Tried a while back with no joy. Speedfit rep couldn't understand why not since they do the 32mm MDPE (?) in large coils of course.

[Mike]

I turned off the upstairs system and with the Alpha near max on the variable pot settings, the return temperature went up to 76 deg. So it would appear the pipe cannot source upstairs and downstairs so I'll need another parallel pump and set of pipes from the boiler to upstairs.

However, even with the return temp correct, the thermostat still has a huge hysteresis. It turns off at about 90 deg C but doesn't restart until the output temperature drops to about 75 degC. At this point the input temperature is still the same, i.e. a degree higher going into the boiler than coming out, presumably due to the large thermal mass of the water in the system.

Grant are "on holiday" until the New Year. Any idea what hysteresis an oil boiler control thermostat should have ?

Thanks

Mike

[John]

You will need to ask Grant for this info but some boiler stats have a small differential adjustment screw concealed about the body somewhere. I'd try to be certain that the flow rates through the boiler are adequate before upsetting the factory adjustment of this though

I'd try turning all your stats full up so that all rads are operating together before balancing the system and then go for a lower differential across them say about 10/11 degrees C. As zones close down on the stats the pump head will increase marginally (circulators are centrifugal) subject to the response of the Alpha. Increased head will give a slightly faster flow which in turn will reduce the differential tamp across the rads but not by much. The reduced Dt will give a higher mean rad temp which will be to your advantage

Sounds good - do you live near a fast flowing stream? I have a hankering to move somewhere hilly with a rushing torrent when I retire and do something similar. If the water is pure I'll install a sandfilter and UV unit to become more or less self sufficient energy and water wise at least

[Mike]

They're "on holiday" until Jan 4th :-(

Yeah - we have a farm in the Peak District. There is a spring at the top of the hill which needs a 10micron, 0.5 micron, then UV filter to make it anything like safe to drink. But we could probably put it in a bottle as is and sell it as mineral water :-)

As for stream, we have a good flowing stream running along the side of us. Total drop on our land is >10m but that would rather a major dam so will try to achieve about 5 feet which will yield around 2 to 3kW for the hot water and charging batteries when the mains goes (every few weeks).

Hope you find something similar.

[John]

SNIP . Have traced the problem to the boiler

A thought occurs to me regarding temperature gradients through the boiler shell. Most boilers have tappings for flow and return at either side of the boiler. Some boilers are not happy if flow and return are connected to the same side of the boiler as it "can" lead to still water (or at least slow to move around) conditions surrounding the thermostat phial. What piping arrangement do you have and does it agree with reccomendations in the Installation and /service booklet for your boiler? I would expect a replacement thermostat to be in the region of £20.00 or so unless its a special. Most boilers seem to use a Ranco model stat

[Mike]

I did wonder this. There are tappings each side but it does specifically mention it is OK to use just one side if you are in a corner. But the thermostat pocket is in the top corner away from the tappings being used.

Only had it a few months so if it's dead I want a new one from them :-)

Many thanks

[John]

SNIP

I would not be surprised if the temperature swings are related to just this point. Is it going to be difficult to route the flow pipe over the top of the boiler into the opposite corner so as to create some water movement around the phial? Have you tried taking temperature readings of the boiler around the thermostat phial to see if there is a "dead spot" giving rise to this lagging response Also, it may be possible if your boiler has two pockets or phials to relocate the sensor to the other pocket. Some boilers have two seperate capillaries (control and overheat) with pockets at opposite sides of the water jacket.

Good point! However you could also withdraw the unit from the phial and with pan of water, hob, thermometer and a multimeter check the actual response of the stat out of the boiler. Having more info available will speed any reponse you do get from Grant

[Mike]

specifically

Thanks for this suggestion which I've just done. Unit switches off at

80degC and back on at around 65 deg. I assume therefore there is a 10 deg drop in the pocket but there still appears to be a huge hysteresis outside of the boiler.

The boiler itself appears well insulated but I couldn't find more than a degree or two difference side to side.

Roll on Grant coming back on Jan 4th.