Balancing and delta T with a Modulating Pump

[Little Tinker]

OK I see what you're getting at. The boiler relay fires for both hot water and heating demand, so if you configure Evohome for S-Plan with only a boiler relay and hot water relay (no heating zone relay)...

Yes, that's it. But I' not sure I'd call that S-Plan anymore. If that is what you have to tell EvoHome then fair enough, but I'd have thought it would understand the setup without having to call it that.

What you are proposing is functionally equivalent to simply removing the heating zone valve (or wiring it always on) and connecting the boiler relay directly to the boiler and having just a hot water zone valve

Yes, agreed. I see the outcome as closer (in terms of final function and what the controller needs to know about) to removing that valve than I do to S-Plan. One of my points upthread was that I don't think the Synchron motors (if that is what are still used...) like being constantly powered, hence why what I proposed may be a better solution as it cycles the power. I thought you could get blanking plates to go on the Honeywell valves in place of the valve "ball bung" that could simplify "removal" for those that were looking for an easy option. Like this part of the conversion kit, but just a flat plate:



but can't track them down at the moment. Not too hard to make though.

However in your suggested configuration ... I think you'd find this would cause temporary room temperature overshoots when hot water demand occurs...

I don't disagree, but suspect that would be a relatively rare and not highly noticeable effect. Presumably you get the same if building a system from scratch with just HR92 and HW zone valve?

Make sense ?

It always did to me . I think we're now agreed that you can have boiler cycle control with just two relays, which helps me feel I understand my options better. Many thanks for that .

Whether to pay the extra £40 for a third relay and change plumbing / wiring for ultimate control and avoid possible oddites remains an option that I'll consider now that I understand it better.

[DBMandrake]

Yes, that's it. But I' not sure I'd call that S-Plan anymore. If that is what you have to tell EvoHome then fair enough, but I'd have thought it would understand the setup without having to call it that.

Well the Evohome doesn't call it S-Plan in the menus it just calls it "Stored hot water control", so it's just me that keeps referring to it as S-Plan. You could even use the hot water relay to control an electric hot water heater (via a suitably rated contactor) as long as the boiler was fired by the hot water relay/zone valve and not a boiler relay. (To avoid the boiler firing when only electric hot water is being called for)

With one, two and three relay configurations possible the system is quite flexible in how it can be configured.

Yes, agreed. I see the outcome as closer (in terms of final function and what the controller needs to know about) to removing that valve than I do to S-Plan. One of my points upthread was that I don't think the Synchron motors (if that is what are still used...) like being constantly powered, hence why what I proposed may be a better solution as it cycles the power. I thought you could get blanking plates to go on the Honeywell valves in place of the valve "ball bung" that could simplify "removal" for those that were looking for an easy option. Like this part of the conversion kit, but just a flat plate:



but can't track them down at the moment. Not too hard to make though.

If you want to bypass the heating zone valve without removing it an easy way to do it is to wire the power for the zone valve across the pump. That way it will always be open when the pump is running, but will not remain permanently open wasting power when the heating is off, or potentially seizing from lack of movement.

This comes up quite often because many people just disconnect the heating zone valve and put it on the manual fill position without realising that opens the valve less than half way. (And the lack of movement might cause it to eventually seize up)

[Little Tinker]

...and put it on the manual fill position without realising that opens the valve less than half way.

That comment got me thinking. I grabbed a spare 4043 Honeywell valve that I happen to have and did a bit of measuring.

As can be inferred from the conversion kit picture above the valve works by moving a ball shaped stopper into a same radius concave surface, which has a hole in it to allow flow from outside the valve. It's similar to lowering an ice-cream cone into a round funnel (where both have the same angle of surface). When the stopper is resting against the housing no flow occurs. As the stopper moves outwards a gap forms between the housing and ball (or funnel and cone). There comes a point where moving the stopper out has no appreciable effect on flow, since whether there is a "massive" or "big" gap is irrelevant compared with the restriction that the narrow outlet part of the funnel creates. The entry hole area becomes the main restriction, not the stopper location.

My measurements were a bit rough, but I got 14mm for the hole into the valve chamber, 4mm for the latched stopper displacement, and 8mm for the full travel displacement (this is by pushing the pivot arm fully open under the motor - it may open less when powered).

If we took the stopper out completely the main restriction would be the entry hole, which has a cross section of:

• Hole: PI*(14/2)^2 = ~154mm²

As the ball opens we can approximate the flow area it creates as being the curved surface of a cylinder with the same diameter as the hole feeding it, and a length equal to the displacement from closed. Crunching the numbers for the two valve positions:

• Latched: 2*PI*(14/2)*4 = ~178mm²
• Full: 2*PI*(14/2)*8 = ~352mm²

So while its certainly true that the actual travel distance when latched is half that of full travel, the flow area it creates is similar to the entry hole size (within my measuring errors). My intuition is that latched is therefore more than half open in terms of flow restriction. I'd expect that the way fluids flow would mean there is some turbulence / non-laminar flow around the stopper that makes it quieter and less restricted when more open (especially at higher pumping rates) and it'd be interesting to see what the practical effect of that is. Is it shown anywhere?

[DBMandrake]

Interesting that you went to the trouble to work this all out.

I didn't take the valve right apart but I think you'll find that the ball doesn't seal against a spherical concave surface, it just seals against the end of a cylinder formed by the flow pipe - if you look at the ball after it opens there is an obvious thin circular imprint left on the side of the ball from the end of the tube that it seals against. Minimising the contact area by having a sphere plugging a cylinder would actually give a better seal because all the force of the spring pushing the ball against the seat would be concentrated on the smaller contact point between the two and actually compress the rubber slightly where it touches, a bit like how a high pressure o-ring seal works.

It also means you'd get a perfect seal without requiring a precise match in diameter between the cylinder and sphere as provided the sphere is big enough it will intersect perfectly with a smaller circle formed by the end of the tube. If you tried to seal the sphere in a spherical curve if they didn't match perfectly you'd get a poor seal and the force being distributed across too great an area wouldn't give a good seal.

Regarding the area calculations I would point out that the restriction provided by cross sectional area is cumulative. So even if the cross sectional area around the edge of the ball was the same as that of the pipe feeding it that doesn't mean no further restriction is added. Far from it.

If that were the case, a 10 metre length of 15mm pipe would give the same flow restriction as a 1 metre length of 15mm pipe - which is not the case. Restriction is both cross sectional area and length. Using an electrical analogy pressure is voltage, flow is current and resistance is flow restriction.

A particular length and diameter of pipe gives a certain flow restriction - the resistance. If you add another identical piece of pipe after it (double the length) you increase the flow resistance - this is the same as connecting two resistors in series - the net result is double the resistance. With water in pipes it doesn't double because flow resistance is not linear (since it is caused by turbulence) but it does increase quite a bit.

The way I tested the change in restriction was to open all radiators fully then determine the differential pressure by twiddling the ABV to find out where the opening point is (you hear it start to flow quite clearly when it starts to open) and compared the valve on the fill position with the valve in the powered on position and I found that the differential pressure across the ABV was quite a bit higher with the valve half open on the fill position vs the powered open position.

In fact the difference was enough that I could hear the motor speed change quite a bit which is an indication of the change in flow and pressure. Before I did any of this I had noticed that after I fitted the zone valve and temporarily locked it open the radiators were heating quite a bit slower than they used to with no zone valve. I did this after an S-Plan conversion where I needed heat for the evening but hadn't had time to do all the wiring changes yet, so I just locked the valve open.

I remembered that the valve only opened half way on the manual position so I quickly wired it to the pump supply for the night and the heat up time of the radiators went back to normal.

So it's not a deal breaker but the difference in performance was quite noticeable on my system. I already have high flow resistance due to microbore piping on most radiators, I don't need to add any extra resistance!

[Little Tinker]

Interesting that you went to the trouble to work this all out.

I find it a very interesting problem .

I didn't take the valve right apart but I think you'll find that the ball doesn't seal against a spherical concave surface...

You're right of course - I was thin king of something similar I'd been playing with recently.

Regarding the area calculations I would point out that the restriction provided by cross sectional area is cumulative. So even if the cross sectional area around the edge of the ball was the same as that of the pipe feeding it that doesn't mean no further restriction is added.

Agreed. I was thinking that at the latched position it will have a similar resistance to a few cm of pipe, but that was somewhat daft since the entry hole is closer to 15mm pipe cross section and a length of that in the main flow line would not be a good thing. It's probably better to think about what creates the equivalent flow area to 22mm pipe, and the ~340mm² of that is pretty close to what the valve creates at the full travel point. Almost as if it was designed that way...

The way I tested the change in restriction was to open all radiators fully then determine the differential pressure by twiddling the ABV ...and I found that the differential pressure across the ABV was quite a bit higher with the valve half open on the fill position vs the powered open position.

Even given the above rethinking it's interesting that it makes such a difference.

[richardc1983]

The pump setting can only be changed from min modulation to max modulation. There are no fixed speed settings.

I started with max modulation thinking this would allow the pump to reduce flow more but basically for the 1 to 2 minute burner on cycles it runs at 100 percent. When the burner turns off it then goes into a 3 minute pump overrun where it drops to 85 percent.I monitored this for 3 or 4 on/off cycles and it was identical each time. I didn't look at individual rad Temps but at the boiler, with a flow temp of 55 degrees, I was getting a retern of approx 42 to 45 degrees.

I then tried changing the pump setting to min modulation to see what affect it has. Now the pump will initially run at 100 percent for approx 10s as the burner kicks in then drop back gradually to approx 60 percent for the remainder of the burn before going to 85 percent for the overrun period. The return temperature being a couple of degrees lower. During this phase of heating (tpi cycling after initial warm up) these setting seem pretty good with pump modulating, a descent delta T at the boiler and we'll into condensing return temperatures.

My plan is now to leaves everything as it is and see what it does from cold with all the rads drawing heat. Fingers crossed for a similar result.

Strange that 'max modulation' does not seem to modulate at all in these circumstances but 'min modulation' does. The manual states that min modulation is best for reliability whilst max is for well balanced systems and can save energy running the pump. That description doesn't seem to fit what I am seeing at all. Maybe there is something else going on with the boiler control logic which I'm missing.

I agree with you here, I have the Lyric T6 opentherm thermostat and have just had the vogue installed. "min" setting seems to give maximum modulation but it never drops below 50% which I think it should, at 50% I get a 9C drop between flow and return. My boiler still prefers to cycle off though and doesn't seem to want to modulate the flow down far enough instead the room thermostat ends up satisfied and the boiler will display "no demand". I have written a separate post about my problems, Ideal have been useless and you end up speaking to a call centre person with less knowledge than me.

http://www.automatedhome.co.uk/vbull...ight=opentherm

[richardc1983]

As we have two upstairs towel rails - the one in our ensuite constantly on as it is a manual valve plus a landing radiator always on, I suppose I could almost shutdown the ABV or leave it set on minimum.

The ABV is a Honeywell DU144 type which sets from .1 bar to .6 bar so I presume I should probably now set this at say .1 or .2 bar at most so it is almost closed.

Am I correct on this setting or should it be .5 or.6 bar for almost closed?

I presume I could temporarily close it fully and see what happens?

Did you have any luck with the adjusting the automatic built in bypass? If it is the one you mentioned above then the boiler requires a minimum flow of 1116 litres/h @ 20c differential and @ 2.5m head. Looking on the chart below then the bypass should be set at 0? The factory set is 0.2 bar if you buy it new so you would think this would still be set wrongly.

http://www.honeywelluk.com/documents...de/pdf/950.pdf

I have 7 radiators in total its a 2 up 2 down. The system is balanced, smallest rad is half of 1/4 turn open and gets hot still. The furthest rad is open a full turn. Large double convector rad is half a turn open and furthest rad is 2 full turns. However it seems to make no difference if they are all fully open or as I have mentioned in the previous sentence. I see a 9c drop across the flow and return. I read that its important to have it balance but is this then causing water to short circulate in the boiler on the auto bypass, particularly when the bedroom TRVs close off due to keeping them at a lower temp. This then causes the boiler to cycle. Then you have more flow restriction because then you will have some rads even though calling for heat with only a fraction of water going through. In fact there is only the bathroom radiator which is a modern fancy vertical radiator (not a towel rail) that is fully open on the balancing valves - no TRV on this radiator. The other rad that is 2 full turns is always turned up full as its quite a small rad for the room. Its a case of do you or dont you, if you have all balancing valves fully open and rads dont clear enough heat then your going to have hot water that's not lost enough heat going back to the boiler this way or if you balance them down then the ABV will then open up doing the same.

So the question is to balance or not to balance, plumber said no point balancing as its a small system but I thought he was fobbing me off as this can be fiddly. When I note the boiler burner cycles off the pump speeds up from 50 to 85% and I note the flow and return temps become the same.

[DBMandrake]

Did you have any luck with the adjusting the automatic built in bypass? If it is the one you mentioned above then the boiler requires a minimum flow of 1116 litres/h @ 20c differential and @ 2.5m head. Looking on the chart below then the bypass should be set at 0? The factory set is 0.1 bar if you buy it new so you would think this would still be set wrongly.

Are you sure the built in bypass valve is set to 0.1 bars from the factory ? I find that very unlikely as this is not enough differential pressure for most systems to get a reasonable flow through the radiators. 0.3 to 0.5 bars would be a more normal setting although it does of course depend on many factors in the system.

[richardc1983]

Are you sure the built in bypass valve is set to 0.1 bars from the factory ? I find that very unlikely as this is not enough differential pressure for most systems to get a reasonable flow through the radiators. 0.3 to 0.5 bars would be a more normal setting although it does of course depend on many factors in the system.

According to http://www.honeywelluk.com/documents...de/pdf/950.pdf 2nd page it shows factory setting in the diagram.

If you work out what the minimum flow the Vogue 32 needs its requires a minimum flow of 1116 litres/h @ 20c differential and @ 2.5m head. Looking on the chart in the above link then the bypass should be set at 0? The factory set is 0.2 bar if you buy it new so you would think this would still be set wrongly.

[DBMandrake]

The link you posted shows the ABV defaulting to 0.2 bars not 0.1. In any case you were originally referring to a built in bypass in the boiler itself - I was saying I doubt that was set to 0.1 bars, if it has a safety bypass internally it will be set quite high and only open under adverse conditions to protect the boiler.

As for the graph you're referring to - without knowing what speed the pump is set to and therefore what its maximum head is and what its load lines look like (pressure vs volume flow for the given speed setting) then taking one point on the graph as you have is a bit meaningless.

Achieving minimum flow for the boiler when you have an ABV is mostly about the adjustment of the pump not the ABV - when all radiators are closed all the flow will go through the ABV regardless of whether its set to 0.1 or 0.6 bars, the only difference is that at a higher setting the higher differential pressure will put a bit more load on the pump and slow down the flow somewhat, but not by a huge amount. That's where the load lines for the pump will help you.

Meeting the boiler minimum flow is not the primary criterion for setting the ABV pressure, you do that with the pump speed and there are a number of other conflicting requirements that dictate the optimal ABV setting, but its almost never going to be 0.1 bars unless you had a really low resistance pipe network in your house.

Some of my radiators are microbore (which does of course have more flow resistance than 15mm) and even with all radiators in the house turned fully on the differential pressure is still about 0.3 bar even with no flow through the ABV. If I were to turn the ABV down to 0.1 bars it would drastically reduce the flow through the radiators and send most of the flow through the ABV and straight back to the boiler. (I set my ABV to about 0.4 bars after weighing up conflicting requirements)

Ideally the ABV should not be flowing at all if at least a couple of radiators are fully open, but I can't achieve that on my system due to the flow resistance of microbore. With 15mm radiator feeds around the house it should be possible though.