Evohome with Heatpump - Shortcycling Inefficiency due to limited boiler settings.

[ecoheated]

Running a fixed speed heat pump from Kensa. They have high starting currents and poor efficiency below 6 minutes.

There is a very good report here on how inefficient it is to run the heat pump for < 10 minutes.
https://assets.publishing.service.go...heat_pumps.pdf

So with the Evohome, i have set the maximum # of starts per hour to 3. Which is fine doing a good job of limiting how many starts.

Its the minimum run time to the heatpump(boiler) that is limiting my efficiency. The maximum allowed is 5 minutes. I would like to set it for 10 minutes to eliminate this shortcycling when all rooms are close to target (Most of the time at start/end of winter season).

With underfloor heating in all rooms and a GSHP, 5 minute cycles is just too short. These systems need less but longer cycles to be efficient.
These systems have expensive buffer tanks added to make sure the heatpump can run for its designed minimum period, but evohome is cutting it short, making most of that redundant.

Any ideas how i can work around this limiting 5 minute on-time ?

Example of 2 hours at low load (all rooms close to target). Evohome uses all of it maximum 3 x start per hour, and its maximum 5 mins per start. So 15 mins / hour, or 25% run time.
It needs more flexibility: 1 x 10 minute run or even 1 x 15 minute run. My hands are tied by Evohome and its frustrating.

Capture12.JPG

[DBMandrake]

I don't think Evohome is designed for controlling a heat pump system, so you are probably trying to hammer a round peg into a square hole.

The problem with trying to set the minimum on time to 10 minutes in a 20 minute cycle is that any heat demand that is less than 50% (10 minutes out of 20) is NOT going to trigger your heat pump on at all, as the minimum on time also acts as a demand threshold.

This would totally wreck the proportional control of the system under a light load, and probably result in rooms dipping well below set points before the heat pump comes on. (Probably why it doesn't let you set the minimum on time this high...)

I've discussed this high minimum on-time issue in another thread recently in relation to using 12 cycles an hour where the 1 minute minimum on time corresponds to a 20% heat demand, and even that is problematic in low heat demand situations in my testing.

If you're totally set on having 10 minute minimum on-time I have another suggestion for you though that might work, put a pump overrun timer between the BDR91 and your heat pump. I use the following timer on my system as an external pump overrun timer, and it would be suitable configured in resetting mode:

https://www.cpelectronics.co.uk/ener...mers/mrt16-rem

In resetting mode as soon as the BDR91 switched on the timer will switch on the heat pump. When the BDR91 switches off the timer will start counting and keep the heat pump on for an additional X minutes before going off. Effectively it will "stretch" the minimum on time of each cycle. The delay time can be adjusted all the way from a few seconds to about 2 hours. (I have mine set to 12 minutes to be slightly longer than the default 10 minute TPI cycle so my pump doesn't stop unless there is zero demand)

So you could achieve your 10 minute minimum on time by setting Evohome to 5 minutes and the timer to 5 minutes, or you could set Evohome to 1 minute and the timer to 9 minutes etc.... While the minimum run time would be the same in both cases the overall system response to changes in heat demand would be different because in the first case you would need a 25% heat demand (5 minutes out of 20) before the heat pump was turned on at all, while in the latter case you would only need a 5% heat demand (1 minute out of 20) before the heat pump would be activated, for a minimum of 10 minutes.

I would say that you may want to do a bit of experimental tuning to adjust the balance of minimum on time between Evohome and the timer, and I would probably start with a 2 minute minimum on time in the Evohome config and an 8 minute delay on the timer, as that will cause your minimum heat demand threshold to be 10%, which is the same as the default one minute minimum on time 6 cycle scenario. If necessary increase the Evohome setting and reduce the timer delay a bit.

One drawback of the overrun timer is that when Evohome commands the heat pump off (heating off action) it would continue to run until the timer expired instead of immediately going off.

Does your system use a heating and/or hot water zone valve controlled by Evohome as well or just the one "boiler control" BDR91 ?

[ecoheated]

I don't think Evohome is designed for controlling a heat pump system, so you are probably trying to hammer a round peg into a square hole.

The problem with trying to set the minimum on time to 10 minutes in a 20 minute cycle is that any heat demand that is less than 50% (10 minutes out of 20) is NOT going to trigger your heat pump on at all, as the minimum on time also acts as a demand threshold.

Thank you for the detailed reply! I found your suggestion of a single shot timer interesting as it was the only solution i had in mind if i cant fix the Evohome setup (which seemed unlikely).

I think i have also come to the conclusion that Evohome is not well suited to Heatpump systems with its design to pulse the boiler so often.

I do have a question following your comments above.
Is this demand % rule that triggers the boiler also true for the underfloor heating HCC80R controllers? I ask because the page that shows demand % for each room on the main control panel doesn't work for the underfloor heating controllers and they seem to have a different demand logic in the background?

I will share photos of what i mean when i get home later.....

Thanks!

[DBMandrake]

I do have a question following your comments above.
Is this demand % rule that triggers the boiler also true for the underfloor heating HCC80R controllers? I ask because the page that shows demand % for each room on the main control panel doesn't work for the underfloor heating controllers and they seem to have a different demand logic in the background?

Unfortunately I have no experience with the HCC80R or underfloor heating, so I don't know the answer to that question.

What I do know is that when you use a boiler control relay with TPI if the heat demand would result in a lower on-time than the configured minimum on-time, the relay simply won't come on at all.

So if you set it to a minimum of 2 minutes on a 10 minute cycle and the heat demand (as reported by the system installer summary page) was currently 15%, the boiler would never come on, at least not until the heat demand crept up past 20%, (as the rooms fell even further below their set points) then it would come on for 2 minutes out of 10.

My personal opinion is that the way it should have been designed is that if the demand was non-zero but below the minimum on time that it should just run for the minimum on time. So in the scenario above, a 0% demand means it stays off, 1-20% means it runs for 2 minutes and then above 20% it is proportional. That would have solved the issue of zones falling below their set points until enough demand was built up to fire the boiler. Unfortunately that's not the way it works.

This issue doesn't apply to OpenTherm control - Any non-zero heat demand will send a "boiler on" signal via Opentherm and the requested flow temperature will be proportional to the heat demand, and it is then up to the boiler itself to decide whether it needs to fire or not to meet the requested flow temperature. This is one reason I'd like to move to Opentherm control in the future.

[ecoheated]

Opentherm looks like it does solve this issue!

I will report back later on some testing i will do this evening to see how the underfloor demand is handled.

[apkent]

I'd love to make these cycle/minimum on time changes to my system but it's bound for HW cylinder control and so I'm not presented with the options in system parameters. Anyone know a way round this? I can't understand why it wouldn't be applicable....

[paulockenden]

You only get those options if you have direct boiler control.

If you have an S or Y plan system wired in the 'traditional' manner (i.e. with the boiler being controlled by the microswitches in the zone valves) then those parameters are hidden.

I've no idea why!

P.

[apkent]

Madness. Yes, we're S plan.

[ecoheated]

Quick update.

Settings:
Maximum cycles per hour: 3
Minimum run time: 5 mins

All zones off.
Then, turn on one zone at just 0.5 degC above current temperature of room.

Interesting results over the 90 minute test. The demand continues to increase as the room is not warming up (i chose the room with the slowest response under-floor heating), which is as expected / hoped from TPI.

Quick observation on the underfloor heating controllers.
When viewing the "System Summary" page in settings, you will never get the demand for each room when using the underfloor heating controllers.
Tonight i figured out that the total demand for all the rooms on the controller is displayed under the room set to output #1 on the UFH controller.
I also noted that the boiler demand is always equal to the highest demand shown from any of the UFH controllers.

So for example. 1 of my UFH controllers does 5 rooms. I set 4 rooms to well below current temp and set the Study to 0.5 above current temp.
The "System summary" page displayed heat demand for the Lounge of 25% and the boiler as 25% too. The lounge is output #1 on the UFH controller.
Later when i had all rooms demanding heat, the Lounge displayed 88% and the boiler demand was 88%.

Anyway, the results are below.
Looks like the UFH pump (which is also when the zone valve opens) is not in sync with the boiler output at all. What i dont like is at 42 minutes the boiler is turned on and the UFH is off, so the boiler is working against no load at all. Not good at all!
And then you also see the UFH running with the boiler off? Again pointless and bad control.

Interest is that even with just 1 of 8 rooms demanding heat, and that room being within 0.5 degC of target, the demand is 25% and increasing.
I think this is a good result as it appears that even with very small loads, the demand for the boiler is kept above minimum start threshold.
I repeated this same test with another room and it came straight in at 28% and the boiler fired.
This morning i noticed that the boiler demand was just 20% and the boiler (heatpump) was not being fired, so that confirms the earlier poster's comment. However all rooms were at or above target, so that was reasonable!

Note: The bottom two plots dont refer to the Y Axis, they are simply on/off plots.
Capture.jpg

[Somebody]

Note: The bottom two plots dont refer to the Y Axis, they are simply on/off plots.
Capture.jpg

How are you extracting the data for the plots please? Are they direct from Evohome? I can chart zone temps and changes from idle to heating via InfluxDB in Smartthings. I would like to track Boiler on off as you've done.

Ta