Need help understanding Honeywell T4 thermostat optimisation

[Ric2013]

I've tried to email Honeywell, but their web address is playing up and they aren't receiving. Unfortunately Honeywell is extremely poor at providing explanations even when I found the firm's online explanatory PDF that tried to explain what Honeywell means.

Anyone had an answer that answers the email I have tried to send Honeywell below:

Dear Sirs/Mesdames,

I need to explain delayed start optimisation on a T4 thermostat to a
homeowner but do not fully understand it myself in spite of reading your
website's PDF.

Optimum start is easy to explain in that the user just needs to remember
that the set start time will be the time the room reaches the set
temperature, not the time the system will fire.

As far as I can follow you, you say the delayed start setting is the
earliest time the system will fire. If this were in ADDITION to optimum
start, it would make senseto me as I could, in effect, tell the T4:


get the house warm by 8, unless that involves firing before 6 - in which
case, fire at 6 and what will be will be



but as delayed start is INSTEAD of optimum start, I can't say I
understand how your thermostat decides what time to fire the system.
Except that the warmer the room is, the later the start. So my
understanding of delayed start is really muddly and is limited to:


[Start setting 6.00] start firing at 6, unless the house isn't that cold
(relative to the temperature required or in absolute terms???) in which
case do what you think



which is obviously nonsense as the thermostat must have an objective way
of quantifying 'isn't that cold' and, being a computer, cannot think and
must therefore be following some kind of formula or IF.. .THEN function.

Can you please help me understand what orders a T4, in delayed start
mode, is following as I need to remain more knowledgable than my
customers?

Many thanks.

Yours faithfully,

 

[Deleted member 120897]

Not something I’ve dealt with personally, so I could be wrong, but the way I interpret that, is similar to how you do, in that it will delay firing the boiler if the room is already say 21 degrees and the set point is say 20 degrees. So it may start on low fire because it doesn‘t need much energy wasted to heat up, if at all.

 

[Chuck]

I've tried to email Honeywell, but their web address is playing up and they aren't receiving. Unfortunately Honeywell is extremely poor at providing explanations even when I found the firm's online explanatory PDF that tried to explain what Honeywell means.

Anyone had an answer that answers the email I have tried to send Honeywell below:

I haven't got a T4 to hand to check but this is my guess:

My understanding is that 'optimum start' only applies to the first programmed temperature in the day, 'optimum stop' only applies to the last programmed temperature in the day. I think they probably originated as features in commercial building controllers where you want the workplace to be up to temperature at the start of the working day and shutdown at the end.

Domestic users, however, are used to time-switches that turn the heating on at a particular time in morning, say 0630 chosen for comfort in cold weather. The 'delayed start' feature compensates for the fact that if it's been a relatively warm night 0630 is necessarily early.

So, I think you are expected to use either 'optimum start' or 'delayed start' but not both at the same time.

Call me a cynic, but I suspect that these features are mainly included to get brownie points on energy efficiency score cards and are of used in practice by very few owners.

 

[John.g]

I think these type of smart thermostats learn over a period of time how long it takes to get the room up to setpoint from different initial (morning) room temperatures and then fires the boiler to have the room up to temperature by the programmed time, and shuts the boiler off a little early for programmed off time (perhaps at the last programmed off time, as suggested above).
It says above that this optisimation is selectable on this stat.

 

[Ric2013]

Not something I’ve dealt with personally, so I could be wrong, but the way I interpret that, is similar to how you do, in that it will delay firing the boiler if the room is already say 21 degrees and the set point is say 20 degrees. So it may start on low fire because it doesn‘t need much energy wasted to heat up, if at all.

It's a 2 wire stat. It can't tell the boiler to low-fire. Unfortunately, it's not my interpretation: that is literally how vague Honeywell's technical explanation PDF is. But we don't know if it delays by half an hour if the difference is only 1 degree or 4 hours. I'll post a link to the document!

I haven't got a T4 to hand to check:

I have, and the installation and user instructions, but they don't help at all

My understanding is that 'optimum start' only applies to the first programmed temperature in the day, 'optimum stop' only applies to the last programmed temperature in the day.

That's an interesting idea, though not supported by Honeywell's literature. Which doesn't mean it isn't correct, though.

Domestic users, however, are used to time-switches that turn the heating on at a particular time in morning, say 0630 chosen for comfort in cold weather. The 'delayed start' feature compensates for the fact that if it's been a relatively warm night 0630 is necessarily early.

It has occurred to me Delayed Start could be a concession thrown in for people who just can't get their heads around the idea of telling the system what time they want the house warm by.

So, I think you are expected to use either 'optimum start' or 'delayed start' but not both at the same time.

You literally can't select both options, as per my email to Honeywell. If you could, if would make more sense!

Call me a cynic, but I suspect that these features are mainly included to get brownie points on energy efficiency score cards and are of used in practice by very few owners.

Absolutely! But they'll never be used if we don't understand them. I have to hold my hands up here though. Even in my own house, I use the programmer to set the on time in the morning because I want warm radiators when I'm having breakfast (even if the house remains cold - so the thermostat and the warmup times are both irrelevant). And I still have a 1980s mechanical stat that I can very easily tweak up or down down depending on my own perception of the temperature when I'm at home in the evenings.

I think these type of smart thermostats learn over a period of time how long it takes to get the room up to setpoint from different initial (morning) room temperatures and then fires the boiler to have the room up to temperature by the programmed time.

Exactly, but what you have described is what Honeywell calls its Optimum Start, which Honeywell has already explained in detail, and not its Delayed Start feature which it really hasn't. Some manufaturers regard these terms as interchangeable: Honeywell doesn't.

This is the Honeywell explainer PDF which, you'll see explains Delayed Start in very vague terms: https://youlearn.honeywellhome.com/uploads/documents/FAQ_-_Optimisation.pdf

 

[John.g]

Just a quick glance seems to tell me that the optimum setting will get the room temperature up to SP by firing the boiler earlier based on historical data but delayed start will never start before its set time but may further delay it again based on historical data, can't see how the two can logically be selected together?

"It's a 2 wire stat. It can't tell the boiler to low-fire." It doesn't have to in a sense as the boiler controls on Flow SP so may low fire or modulate once the SP is reached but I agree that it doesn't specifically tell the boiler to run at a fixed output,10% 20% etc for a fixed time.

 

[Basher]

Just been Googling 'cos I was intrigued by this.
First thing to say, which I hadn't realised, but you probably know, is that the terminology "optimum start" and "delayed start" seem to be universally used across different brands of thermostat. So looking at explanations of other products does slightly help lift the fog.

As I now understand it, invoking "delayed start" causes the thermostat to take a temperature measurement (say) 45 minutes before the set 'on' time, and then decides on the basis of the difference between that 'early' temp measurement, and what the required target temperature is, when to first turn on the boiler.

That is a subtlety different approach to learning how early the boiler needs to be switched on to guarantee being at target temperature by the set 'on' time.

Don't take this as gospel - it might be worth looking at other manufacturers explanations of "delayed start" because this seems to be an established eco 'thing'

 

[Chuck]

That's an interesting idea, though not supported by Honeywell's literature. Which doesn't mean it isn't correct, though.

I didn't just make the information up, my description was based some Honeywell instructions or briefing that I've read in the past but I can't remember exactly where. Possibly an older model of temperature programmer or some training material from one of their courses.

You literally can't select both options, as per my email to Honeywell. If you could, if would make more sense!

Having 'optimum start' and 'delayed start' as mutually exclusive options, makes perfect sense to me.

The Honeywell FAQ you cite seems to be a quite clear explanation to me and matches what I wrote. The summary at the end that says that 'optimum start' optimises comfort possibly at the expense of additional fuel consumption and 'delayed start' reduces fuel consumption but may leave the house a bit chilly in the mornings encapsulates the difference between the strategies nicely.

 

[Ric2013]

Just wanted to clarify I'm not after the difference between the two modes, but specifically the formula by which the T4 thermostat computer calculates the time period by which it delays the firing time after the set time in Delayed Start mode based on the temperature difference.

Just a quick glance seems to tell me that the optimum setting will get the room temperature up to SP by firing the boiler earlier based on historical data but delayed start will never start before its set time but may further delay it again based on historical data, can't see how the two can logically be selected together?

I got the idea from another manufacturer's website. That other manufacturer allows you to set an earliest on time AND an ideally-must-be-warm-by time. So you could tell the programmer that ideally you wanted 18°C by 9.00, but that it mustn't switch on before 7.00 even if not starting before 7.00 means the room isn't warm by 9.00. This makes sense to me because I would understand the criteria by which the computer is delaying the start time - it would be delaying it the point that will allow the house to JUST warm up by a set time, or a bit later as it would be constrained by the earliest on time.

Honeywell, as you say, allows you to select an earliest on time or an ideally-must-be-warm-by time, but not both. So, logically, the T4 computer must work in one of two ways: 1. it allows a maximum (preset by Honeywell rather than user-adjustable) period between the earliest on time and the time the room will reach set temperature and fire the boiler as late as it can to bring the room to as close to set temperature as it can within that period; 2. it applies a delay period after the earliest on time calculated in minutes per degree difference. THIS is the information I am after and which Honeywell has not (to my knowledge) made public.

As I now understand it, invoking "delayed start" causes the thermostat to take a temperature measurement (say) 45 minutes before the set 'on' time, and then decides on the basis of the difference between that 'early' temp measurement, and what the required target temperature is, when to first turn on the boiler.

That is a subtlety different approach to learning how early the boiler needs to be switched on to guarantee being at target temperature by the set 'on' time.

Don't take this as gospel - it might be worth looking at other manufacturers explanations of "delayed start" because this seems to be an established eco 'thing'

What you have said (except the 45-minute idea) is pretty much the Honeywell explanation and I think you're right, but what I'm after from Honeywell is how the computer calculates the on time based on that measured temperature difference. If it's only 1 degree, does it delay start for an hour or for five hours? How about if it's 3 degrees?

I didn't just make the information up, my description was based some Honeywell instructions or briefing that I've read in the past but I can't remember exactly where. Possibly an older model of temperature programmer or some training material from one of their courses.

No worries, pal. I wasn't suggesting you had!

Having 'optimum start' and 'delayed start' as mutually exclusive options, makes perfect sense to me.

The Honeywell FAQ you cite seems to be a quite clear explanation to me and matches what I wrote. The summary at the end that says that 'optimum start' optimises comfort possibly at the expense of additional fuel consumption and 'delayed start' reduces fuel consumption but may leave the house a bit chilly in the mornings encapsulates the difference between the strategies nicely.

It makes sense, and it seems we agree with what our understanding of what the FAQ states. but it leaves me with the question of what the actual actual delay period would apply with Delayed Start option. Esicontrols has a maximum delay of 45 minutes, Drayton, like Honeywell, doesn't state, and Horstmann says Delayed Start is another name for Optimum Start. Clear as mud!

 

[Ric2013]

UPDATE 20/10/2021

Turns out Honeywell doesn't seem to know how the feature works or is refusing to tell me, which begs the question as to why, if Honeywell bought in the software for the T4 from another firm, it hasn't apparently gone to the trouble of testing what the T4 will do in order to be able to at least provide indicative information as to what is and what isn't normal behaviour. You'd have hoped at least they'd have tested its response to a series of actual temperatures and plotted a chart of what it actually did under these conditions.

I finally managed to have an exchange of emails with Honeywell UK. Apparently the Delayed Start feature will delay the start of heating 'with a maximum of 1 hour', depending on the 'learnt heat up range' and the Honeywell adviser went on to suggest Optimum Start is a better feature anyway and needlessly explained how to set it up as if I were incapable of reading an instruction manual.

The adviser sadly ignored my question as to what exactly he meant by the term 'learnt heat up range' and whether the delay is in direct relation to the number of degrees difference, or whether the software was more complicated (i.e. is it based on a Rate of Change calculation or a warm-by-time)? Instead, he told me that if the difference is small, then the delay will be a few minutes longer than if the temperature difference is bigger, which was again stating the obvious and gave me the impression I was dealing with a call-centre robot who was reading off a script and not someone with a technical background.

Realising that either I was writing to a complete idiot or that the Honeywell adviser thought he was writing to a complete idiot, I asked how long the delay would be for 1 degree, 2 degrees, etc, and received the reply 'we don't have a calculator to take into account all the software in the product and offer such detailed answers'.

This has been disappointing, but sadly it comes hot-on-the-heels of a friend telling me the Honeywell wireless receiver I installed for him not that long ago wasn't working and, as I patiently talked him through faultfinding by phone (while in another country), it became obvious the receiver itself had failed (I think the transmitter batteries actually outlasted the receiver). And that wasn't installed long after I had problems with a Honeywell ABV failing to close and Honeywell technical being unable to suggest any remedy except returning it to the merchants (turns out the valve was prone to getting jammed open if there was any loose magnetite in the system and not directly at fault, but it was a pity Honeywell's technical couldn't have told me that and saved me a lot of head-scratching).

I think there is a lesson I need to learn here...

Anyway, thanks to all who contributed. You were far more Helpful than Honeywell was!

 

[Ric2013]

Further update.

I was recently in the flat in which this T4 was installed. The lady who was living there (more of a squatter than anything else - long story) had thankfully moved out and I was carrying out some repair work. I decided to select the optimum start feature to see how slow the flat was to warm up compared with the Honeywell baseline.

As there was some confusion about optimised and delayed start in this thread, I'll just explain that the OPTIMISED START feature is supposed to learn how long the house takes to warm up and select a firing time to allow the room to reach temperature by the set time. This seems a useful feature.

On the FIRST day. The night setback temperature was 13°C and the measured room temperature was 13°C when the heating first started to fire at around 08.30. The room was supposed to be at 18°C by 10.00, by which time it was at 15°C and it only reached 18°C at 12.00. To me this showed that the room was slower to heat than the Honeywell T4 was expecting and, having no historical data to use (as the optimisation features had previously been disabled), it reverted to a guess, which seemed a reasonable approach.

On the SECOND day. The room reached the 18°C setpoint by 11.30. Obviously the thermostat had learned little from its previous day's experience.

On the THIRD day. I was not there when the room reached setpoint, but it was still at 15 or 15.5°C at 10.00, so still nowhere near on time.

On the FOURTH day. I disabled optimised start and set the firing time to 06.30 and it reached setpoint at approximately 10.00.

I am not sure if the thermostat would ever have got the start time right, but either there was a maximum warm-up time and the flat exceeded it, or the programming allowed it to tweak the firing time by a few minutes every day but not make the logical conclusion that if it was reaching setpoint two hours late then it needed to start firing around two hours earlier. A bit like getting to your first day of work two hours late and deciding the solution is to leave home half an hour earlier.

If this pi$$poor [if "Private Eye" can use the word, why does this forum feel a need to censor it?] AI is representative of the AI employed in the heating industry as a whole, then we're going to have trouble. Nothing is ever really going to run as efficiently as it is supposed to unless a human uses actual intelligence to calculate flow temperatures etc and manually controls the system on a day-by-day basis which, at a domestic level, is not realistic. Hopefully Honeywell is the worst and other manufacturers are making far better products!

 

[Lou]

(Just bumping a few older threads, they will drop off what is new soon)

Check Out Plumbing News!