OT produced a heat demand during a comm failure...was that the fail safe?

[bruce_miranda]

Woke up in the middle of the night to hear the boiler running, which was odd because the heating system has been Turned Off for while now. We have solar so even the HW is taken care of.
What I found was one of the zones had the hour glass symbol and that same zone was producing a Heat Demand of 10, as confirmed in the System Settings. Why was that? Was it the fail safe that kicked in?
Worst of all there is absolutely nothing in the fault log to suggest what happened to that zone. No message regarding that zone for months now.

Now when I look at what the various systems were doing its even more puzzling. I've attached a graph to show you what I mean.

The OT Bridge is requesting some really high temperatures from the boiler (thank goodness my boiler is set to a max of 60C in the summer). You can see that the HGI80 manages to read the zone's correct room temperature but the temperature held for that zone on the Honeywell servers is 128C!

KitchenCommsFailure.jpg

All seems to magically return back to normal after about 50 mins or so. Strange.

[DBMandrake]

I assume you are graphing temperatures from the API ? If so, a temperature of "128" degrees actually means "unknown" temperature. You will see 128 at the same time that the controller displays an hour glass. I actually have special case handling for 128 in my own graphing scripts for this reason so it doesn't cause a huge jump in my graphs... Although I don't know for sure, it's possible the controller also sends a temperature of "128" to other devices in the zone when the true temperature is unknown.

Do you have failsafe mode enabled in the controller ? One thing fail safe does is if the controller and heating/boiler BDR91's loose comms with each other for approximately one hour (which will bring on the red LED on the BDR91) the BDR91's will then automatically start cycling on at a 30% duty cycle so your house and pipes don't freeze due to a loss of comms...

However I have wondered aloud on this forum a couple of times whether comms loss between HR92's and Controller would also trigger any fail safe action, and nobody knew. Your graph seems to confirm this is the case - not long after the measured temperature of a zone went to unknown, the flow temperature request increased. And perhaps the brief high temperature requests are to cycle some water through the system to avoid freezing pipes.

In general I leave fail safe mode off during the summer and seasons where there is very low risk of freezing and only turn it on in winter (and especially if I'm away from home) when there is a significant risk of pipes freezing should comms fail.

[bruce_miranda]

The 128C is via the API (honeywell server). But the ~21C is measured by a local HGI80 which shows that the zone valve was still publishing its accurate temperature but for some reason the controller went into an unknown state for that zone. The heat demand that I saw and as confirmed by the HGI80 was 10% (value 20 via the HGI80). What is however the crazy thing is that the OT heat demand being sent to the boiler starts off low but ramps up to 90C I.e. full demand. This was previously known to happen only if the target was below the set point by more than 1C. But in this instance the set point stays put at 5C, I.e. Off.
So what caused the OT heat demand to ramp up? Also the 10% heat demand

[DBMandrake]

The 128C is via the API (honeywell server). But the ~21C is measured by a local HGI80 which shows that the zone valve was still publishing its accurate temperature but for some reason the controller went into an unknown state for that zone.

Just because your HGI80 received a temperature measurement from the zone doesn't mean the controller did. Presumably the HGI80 is not in exactly the same location in the house as the controller so if there is a collision between two different devices in the house then based on proximity the HGI80 might have heard the temperature sensor while the controller heard the other colliding device...

Another factor is how long after not receiving any temperature measurements will Domoticz with HGI80 report unknown temperature ? Or will it just continue to show the last known temperature long past the time when the temperature has last been received ? In the period of time when the controller reported unknown did Domoticz show any variation in the reading or did it just stay exactly the same ? If the latter it may have just been holding onto the last known value.

If the controller reports unknown temperature it's because it hasn't heard a temperature broadcast from the zone for a while.

The heat demand that I saw and as confirmed by the HGI80 was 10% (value 20 via the HGI80). What is however the crazy thing is that the OT heat demand being sent to the boiler starts off low but ramps up to 90C I.e. full demand. This was previously known to happen only if the target was below the set point by more than 1C. But in this instance the set point stays put at 5C, I.e. Off.
So what caused the OT heat demand to ramp up? Also the 10% heat demand

Probably fail safe mode as already discussed. Fail safe mode behaves differently to normal operation. If the controller doesn't know what the temperature of a zone is for an extended period of time it has to assume that it could be freezing. The whole point of fail safe mode is to protect your house against frozen pipes if comms is lost. So as far as I can see, it's doing its job here.

If you don't like the behaviour of fail safe, turn it off - I turn it off except for winter where freezing of pipes is a risk as I don't like the idea of extra gas use due to intermittent comms issues.

The real issue is why the controller didn't receive any broadcasts from the temperature sensor for a long time.

[bruce_miranda]

So you think that the OT Bridge has intelligence built in to determine that a normal 10% Heat Demand from a zone is different from a Fail Safe 10% demand. Because as I mentioned and as can be seen from the graph that the Heat Demand stays flat at 10% but the Target temperature requested from the boiler by the OT Bridge ramps up twice in that period.