Balancing and delta T with a Modulating Pump

[jdp80]

Tacking onto this thread, I'm monitoring boiler flow and return values and do frequently see return higher than flow especially during periods of low demand when TPI is running

e.g.

See the period between 0745 and 0930 when there's marginal demand for a few radiators.

It's a Vaillant 637 so has built-in flow/return temperature sensors which I'm reading via eBus.

[DBMandrake]

Tacking onto this thread, I'm monitoring boiler flow and return values and do frequently see return higher than flow especially during periods of low demand when TPI is running

It's perfectly possible for return to be hotter than flow under certain circumstances and for a short time.

Say some large radiators in the house were heated up to a full flow temperature of 70 degrees with the HR92's open and then they closed. The thermal mass of those radiators would hold that 70 degree panel temperature for quite a while.

After those HR92's closed the heat demand would drop and that might cause the boiler flow temperature to drop if there wasn't much other demand. Say there was another radiator in the house with a small demand that caused the boiler to run at about 55 degrees.

A little bit later the first radiators open again and now because those radiators are hotter than the current boiler heat exchanger temperature those radiators actually end up trying to heat the boiler up via the return pipe (at the expense of cooling themselves down) for a short amount of time causing your return pipe to be hotter than the flow. Heat will flow from anywhere that is hotter to anywhere that is colder, and the big radiators are acting as a thermal store.

[Little Tinker]

You get more control options.

I've decided I really don't understand this . EvoHome can't know what is on the end of the relay. Starting from S-Plan either...

1) You remove the CH zone valve and connect the relay to the boiler. Then whenever EvoHome calls for heat both boiler and pump fire. Water passes any unobstructed path through radiators (as controlled by open HR92 heads), and other paths if they exist.

2) CH zone valve stays in place and is controlled by relay. Then whenever EvoHome calls for heat the valve opens, causing both boiler and pump fire. Water passes through the opened zone valve to any unobstructed path through radiators (as controlled by open HR92 heads), and other paths if they exist.

I understand that the time taken for zone valves to trigger pump / boiler may not be negligible compared with the 1 minute cycle time, but that can't be the rationale can it? What am I missing?

[paulockenden]

With conventional S plan you don't have a boiler relay bound (only ch and dhw relays), with the boiler control coming off the back of the valves. And then you don't get the boiler cycling parameters on the controller.

P.s. Have you tried doing the online training?

[Little Tinker]

With conventional S plan you don't have a boiler relay bound (only ch and dhw relays), with the boiler control coming off the back of the valves. And then you don't get the boiler cycling parameters on the controller.

P.s. Have you tried doing the online training?

I think this would drop into place if I had enough bits to try out different scenarios, or a different house . All I have is controller, HR92 heads, and single BDR91, so my experiences are based on a week of watching those.

I have looked at the online training and installation guides. What I don't see in there is what I assume is a common scenario of S-Plan plumbing but with HR92 heads in place. Maybe this is where I'm getting stuck. My assumption would be that you could set that up as one BDR to HW valve, and one to CH valve with HR92 still controlling temperature in each room. So to get heat to a room controller gas to open rad valve and the zone valve. The guide talks about a multi-room zone, which I think this would be but it isn't clear just from reading whether each of those rooms can then have a separate temperature control, with multiple sensors plus both a zone and room actuator.

I suspect the rationale is slowly dawning though. Whether you can set up that zone valve + HR92 approach or not I thought there was a "work around" of binding a BRD as a boiler controller but connected to the CH zone valve. The weakness I see in this is that the TPI behaviour might cycle the boiler on and off every few mins via the BRD. With "proper" boiler control wiring the pump overrun can then clear heat out of the boiler into the radiators via the open CH circuit. With my approach the CH valve would open and close as the boiler was called, so the boiler internal bypass would have to handle the overrun - which is wasteful, potentially straining to the boiler if done too frequently, and the "missing" heat will probably be confusing to the TPI learning of the controller.

Assuming I have that right I'll have to figure out what to do with my wiring...

[DBMandrake]

I think this would drop into place if I had enough bits to try out different scenarios, or a different house . All I have is controller, HR92 heads, and single BDR91, so my experiences are based on a week of watching those.

I have an S-Plan setup with 3x BDR91 and HR92's on most radiators, but I have also run in a 2x BDR91 configuration during testing so ask away if you're confused on any points.

I have looked at the online training and installation guides. What I don't see in there is what I assume is a common scenario of S-Plan plumbing but with HR92 heads in place. Maybe this is where I'm getting stuck. My assumption would be that you could set that up as one BDR to HW valve, and one to CH valve with HR92 still controlling temperature in each room.

That's exactly how it works. HR92's control individual radiators which form zones. But they also call for heat wirelessly through the controller which will bring on the CH BD91, and also the Boiler control BDR91, if you have one.

So to get heat to a room controller gas to open rad valve and the zone valve. The guide talks about a multi-room zone, which I think this would be but it isn't clear just from reading whether each of those rooms can then have a separate temperature control, with multiple sensors plus both a zone and room actuator.

Multi-room zone is something else entirely.

A standard zone is a "single room zone". In this type of zone there can only ever be one temperature sensor in operation but there can be one or more actuators that act in response to that one temperature reading. So for example if you had a room with two radiators with an HR92 each you would designate one of those HR92's as the temperature sensor (or it could be a wall mounted DTS92, Evotouch controller etc) and the other HR92 would act like a slave, adjusting its radiator but relying on the temperature reading coming in from another device.

When a set point change occurs or if you make a manual override on one HR92, both will be affected. (After a delay)

A multi-room zone lets you group multiple rooms into a single zone that share a single schedule but in a multi-room zone every HR92 uses its own built in sensor instead of relying on a single nominated sensor - a good example would be childrens bedrooms. You could have two separate bedrooms with an HR92 each that follow the same schedule but are able to measure and regulate their own temperature. If you tried to use a single room zone here the room whose sensor was not being used would be at a wildly wrong temperature as its radiator would be getting adjusted in response to a temperature reading taken in a different room. Make sense ?

I suspect the rationale is slowly dawning though. Whether you can set up that zone valve + HR92 approach or not I thought there was a "work around" of binding a BRD as a boiler controller but connected to the CH zone valve.

If you have an S-Plan system but only want evohome to control heating and leave hot water on an existing timer you can do exactly that - have just one BDR91 bound as a boiler control relay but connect that to your heating zone valve and let the zone valve fire the boiler.

Or if you want the Evohome to control hot water as well you would not configure a boiler control relay, you would configure a heating relay and hot water relay and connect those to you zone valves and use the zone valve limit switches to fire the boiler. (Which is standard practice in S-Plan systems)

Yet another option is to have a third BDR91 as a boiler control relay to fire the boiler directly instead of using the limit switches in the zone valves - this is the method I use.

The weakness I see in this is that the TPI behaviour might cycle the boiler on and off every few mins via the BRD.

How is this a weakness ? You want the boiler to be cycled with TPI. It doesn't matter if you control the boiler directly using a boiler control relay or indirectly using a hot water zone valve relay - in both cases they are TPI modulated to the heat demand, and if you have both as I do, they seem to get modulated to the same time intervals.

With "proper" boiler control wiring the pump overrun can then clear heat out of the boiler into the radiators via the open CH circuit.

I'm not quite sure how this has anything to do with pump overrun - whether you fire the boiler directly with a boiler control relay or via a zone valve limit switch, either way its up to the boiler (or an external timer) to provide pump overrun - the Evotouch does not do pump overrun for you. My boiler is very old and didn't originally have any pump overrun so I added an external overrun timer (an MRT16-REM) when I first installed Evohome.

With my approach the CH valve would open and close as the boiler was called, so the boiler internal bypass would have to handle the overrun - which is wasteful, potentially straining to the boiler if done too frequently, and the "missing" heat will probably be confusing to the TPI learning of the controller.

I used to think the same way - until I tried it. Before my S-Plan conversion I had no zone valves and had a long pump overrun timer so during the TPI off period the water continued to flow allowing the flow temperature to drop way down.

Now with S-Plan the heating zone valve is modulated with TPI as well as the boiler relay, so yes, each time TPI cycles the boiler off the zone valve also closes and forces the pump overrun to flow through the automatic bypass valve - which initially I was worried about.

But it has been absolutely fine, and an unexpected benefit I found is that temperature regulation of rooms is actually much better now that the zone valve is in circuit, as it seems to let the system regulate the total heat flow from the boiler to the radiators much more accurately than just cycling the boiler on and off but allowing the pump overrun to continue to flow through the radiators. I also haven't noticed any extra "strain" on the boiler - flow temperature just stays near the flow temperature set point when the zone valve closes.

It certainly doesn't "confuse" TPI - quite the opposite, it seems to give TPI a lot better proportional control. An temperature loss through the ABV loop during the pump overrun in the off part of the TPI cycle seems to be very minimal.

[Little Tinker]

...so ask away if you're confused on any points.

Thanks. The heart of my confusion is why three BDR91 is better than two. I thought I was getting there when considering that boiler overrun has nowhere to go under the relatively frequent TPI control, but your comments suggest that isn't the reason.

This group of sentences highlight the mental barrier that I seem to have:

... if you want the Evohome to control hot water as well you would not configure a boiler control relay, you would configure a heating relay and hot water relay and connect those to you zone valves and use the zone valve limit switches to fire the boiler. (Which is standard practice in S-Plan systems)

Yet another option is to have a third BDR91 as a boiler control relay to fire the boiler directly instead of using the limit switches in the zone valves - this is the method I use.

...

...the heating zone valve is modulated with TPI as well as the boiler relay, so yes, each time TPI cycles the boiler off the zone valve also closes and forces the pump overrun to flow through the automatic bypass valve - which initially I was worried about.

But it has been absolutely fine...

If the CH valve is happy being modulated and the closing of that valve doesn't seem to cause any deadheading, kettling, etc problems for the boiler and pump then why use three relays, rather than two with the CH one set up as boiler control but wired to CH zone valve (which will in turn fire the boiler, and controller can't tell the difference)?

[DBMandrake]

Thanks. The heart of my confusion is why three BDR91 is better than two. I thought I was getting there when considering that boiler overrun has nowhere to go under the relatively frequent TPI control, but your comments suggest that isn't the reason.

This group of sentences highlight the mental barrier that I seem to have:



If the CH valve is happy being modulated and the closing of that valve doesn't seem to cause any deadheading, kettling, etc problems for the boiler and pump then why use three relays, rather than two with the CH one set up as boiler control but wired to CH zone valve (which will in turn fire the boiler, and controller can't tell the difference)?

I didn't say three relays is necessarily better than two, but it might be useful in some circumstances. It gives you the following flexibility:

1) You can use the hot water overrun setting. If you enable this (for example set it to two minutes) then when the hot water reaches temperature it turns off the boiler relay immediately but leaves the hot water zone valve open for an additional two minutes, and assuming you have some pump overrun (which is up to your boiler) this lets the latent heat of the boiler transfer into the cylinder, and also gives the boiler a "softer landing" when it finishes the reheat. If you have only two relays the boiler must be powered by the switches on the zone valve, which makes it impossible to turn the boiler off but keep the zone valve open a bit longer.

2) A 3rd relay gives access to the minimum on time and cycle rate boiler settings in Evohome, which might or might not be useful or necessary for you depending on your system. With only two relays you're stuck at the default 6 cycles per hour and depending on who you believe either 1 minute minimum on time or no minimum on time. This is purely a software limitation in the Evohome firmware - there's no reason in principle that Honeywell couldn't make these two settings available on a two relay configuration as the heating relay does TPI as well, and quite a few people are asking for it, but I wouldn't hold my breath on it happening in the near future. So if you need those settings three relays are your only option at the moment.

3) If you fire your boiler from the switches in the zones valves there is typically a 15 second delay between when the valve starts to open and finishes opening and fires the boiler. This might not sound like a lot but if the system is trying to fire for a 1 minute on time and the boiler doesn't even start until 15 seconds into that minute you've lost all that heating time. With the three relays the boiler will fire at the same instant the zone valves begin opening. The initial few seconds of flow will have to be passed by the ABV - but you need one of those in an S-Plan system anyway. And by the time the valve has fully opened the boiler will be producing useful heat.

4) Maybe your zone valves don't have switches in them, or they are being used for another purpose - for example I use the switch in my hot water zone valve to trigger a flow temperature boost for hot water re-heat.

So there's no major differences between a two and three relay config but depending on your system it might be useful. I suspect most people only use two relays on their S-Plan systems and stick with the original boiler wiring that fires the boiler from the zone valve switches, which is the defacto standard in S-Plan wiring.

If I were to do the hot water kit install again I'm not sure that I would bother with the 3rd relay - I thought hot water overrun might be useful to me but I ended up disabling that, and I wanted to retain access to the minimum on time and cycle rate settings that I had before my S-Plan conversion but I ended up just staying with the default settings anyway!

[Little Tinker]

I didn't say three relays is necessarily better than two, but it might be useful in some circumstances. It gives you the following flexibility:

1) [HW overrun]

2) A 3rd relay gives access to the minimum on time and cycle rate boiler settings in Evohome ... . With only two relays you're stuck at the default ... . This is purely a software limitation ... .

3) [Valve opening delay]

4) [No valve micro-switches]

My comment about "better" was inferred from higher in this thread and others, so I wasn't trying to claim you thought it was .

I agree with 1, mentioned 3 above as one of the possible benefits so agree there (and have notched my minimum run to 2 mins as an attempt to compensate) and see 4 as a reasonably rare case (and different to where I came in on this re Paul's setup where he does have the valve switches). That just leaves point 2, which was the feature I joined this thread in relation to, and specifically trying to understand why three relays are need to access it. I think one more clarification as to how EvoHome actually works will help me see the light.

In my controller (no HW control, just HR92 heads and one relay bound as boiler control but connected to CH zone valve) I can adjust the boiler cycle settings, and am making an assumption that this is because I have a relay bound as boiler control.

On a different thread the following was said:

1) ... you can have a hot water configuration with boiler relay and hot water relay (no heating zone valve relay)...

What I keep looping back to in my head is that setup: two relays bound as HW and boiler relay, plus HR92 heads - but the boiler relay is actually connected to the CH zone valve. As far as EvoHome is concerned you have HW zone valve, boiler relay and HR92 heads, which I expect must be a common configuration and therefore supported. If the assumption that having boiler relay bound is what gives the extra parameters is correct then that would seem to be an answer to Paul's original point re adding 3rd (possibly unnecessary) relay.

Incidentally, this would seem to address point 1 as well as long as the HW orange wire is disconnected. When HW is needed both valves will open, but unless there is an HR92 open no water will pass that on CH path. Contoller can then signal end of call for heat, closing CH valve and hence turning boiler off. But HW valve will allow overrun to pass until it is also closed by controller.

And just to ensure that I'm not seeing different to others because of changes around the software limitation point that you mention, my controller has s/w 01.00.03.00.

[DBMandrake]

What I keep looping back to in my head is that setup: two relays bound as HW and boiler relay, plus HR92 heads - but the boiler relay is actually connected to the CH zone valve. As far as EvoHome is concerned you have HW zone valve, boiler relay and HR92 heads, which I expect must be a common configuration and therefore supported. If the assumption that having boiler relay bound is what gives the extra parameters is correct then that would seem to be an answer to Paul's original point re adding 3rd (possibly unnecessary) relay.

Incidentally, this would seem to address point 1 as well as long as the HW orange wire is disconnected. When HW is needed both valves will open, but unless there is an HR92 open no water will pass that on CH path. Contoller can then signal end of call for heat, closing CH valve and hence turning boiler off. But HW valve will allow overrun to pass until it is also closed by controller.

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) and then connect the boiler relay to your heating zone valve, every time hot water needs to heat up both the heating zone valve and hot water zone valves will open, so you'd be relying on all radiators being HR92 controlled to prevent hot water demand heating up radiators - and its more often the case that people still keep a bypass radiator such as towel rails. (Although I specifically chose to eliminate any sort of bypass radiator on my system - I don't want any radiator I didn't ask for to come on)

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 - again, you are relying on all radiators having HR92's. It's not quite identical though because in a true S-Plan configuration the heating zone valve cycles on and off with the heating TPI duty cycle and this in conjunction with how far open the HR92's have the radiator valves controls the heat flow to radiators. (Both the HR92 valve openings and the heating zone valve duty cycle contribute towards control)

If in this condition a hot water demand came along the boiler relay would now stay on continuously for the duration of the hot water demand but the heating zone valve relay would still cycle on/off under TPI control and maintain the radiator temperatures spot on even though the cylinder was heating at the same time.

However in your suggested configuration the heating zone valve would be TPI controlled when there was no hot water demand but as soon as a hot water demand commenced the boiler relay and therefore your heating zone valve would stay open constantly. I haven't tested it but I think you'd find this would cause temporary room temperature overshoots when hot water demand occurs because the radiators would suddenly heat up a lot more than the system was anticipating, compared to having a properly controlled heating zone valve where it can continue to TPI modulate the valve.

Make sense ?