No low opinion of heating engineers Tony, nor those who chose vocational rather than academic. Its the RGIs (especially the fast-track course/van with name brand on the side type) who think that their gastec cert qualifies them to design heating systems that wind me up.
The big question I had in the original post (and still have; more below) was "to what degree is cycling acceptable/how short is too short?"
Didn't know that about the wall bracket. I'll have to dry fit it then leave it on the floor then, for the RGI to put the screws back in and nip up the compressions.
Have found RGIs willing to replace the existing gas pipe (as the system stands) with a correctly sized one (300 quid) and commission a boiler or two at the same time (add 200/300 quid) which is reasonable for a day's work BUT it looks like I won't get a boiler warranty with that. Approved installer? 2200 for the boiler/500 to fit. Non-approved? 1200/500 to fit. No extended warranty with the guy who charges 500 to fit. Extended warranty with the approved installer, but you're then paying 1500 to fit. At that kind of price you may as well buy the boiler once and buy the parts later as they fail, rather than pre-paying for all the failures upfront. Will look further north/east as there were just the two official Vokera installers in Sutton Coldfield.
Anyhow... ...back to cycling. (expertise required here Breesey; happily confessing what I don't know here...)
You have a house. Say you wish to keep it at 20C +/-0.5C with outside temperatures from 0C to 20C.
If you wanted the boiler to run continuously, then it would need 20:1 modulation/turndown. Say 20kW when 0C outside and 1kW when 19C outside for a house. 100 kW and 5 kW for a commerical application.
The commercial lot with their 100 kW demands using five 20kW boilers with 4:1 modulation/turndown and a boiler management system.
In residential applications the boiler must cycle on and off, as nobody makes a boiler with 20:1 modulation/turndown.
Increasing the hysteresis to +/-1C would drop the modulation/turndown required, but folks notice that temperature swing, and it won't drop it far enough to avoid the need to cycle. I fyou want heating down to -5C the problem is worse.
You can control the cycle length with the water content of the system, hysteresis and setpoints on the flow/return temperatures, and radiator sizing. You can cycle on the room stat, the boiler stat, or indeed both.
What are the acceptable time constants? How short is too short? This is a question a heating engineer (which I am not) will know the answer to.
(At some point the benefits disappear, so "as long as possible" isn't an acceptable answer)
Having done some reading around I'm thinking 10-15 minutes is about as short a time as you'd want the boiler to fire for.
The EU boiler efficiency tests at part load are 10 minutes long and the manufacturers are optimising for these/to cheat these. See PP31 of this document:
http://www.ecoboiler.org/public/ecoboiler_task1_final.pdf
I'm now fairly sure that the "10:1 modulation" Vokera boiler has "cheat mode" built in. The efficiency at 30% load is probably* rubbish, so they deliberately make it fire at 75% (probably the optimum efficiency point...) for the first 15 minutes then modulate down. The EU still gives them an "efficiency at 30% modulation" but it will be from firing for a shorter time at 75% then turning off, rather than the actual efficiency observed at 30% modulation.
You can turn off "cheat mode" in the field using the settings on the boiler as the technical centre described.
On some (Vaillants?) you can't and unless you've got radiators that'll dissipate the full output it'll sit there bouncing off the overheat/anti-cycling control. That'll be an issue with combis with massive outputs, or where radiators have TRVs and the house is close to the setpoint.
SEDBUK is hopeless for choosing between the boilers, as it takes the lab test data then process it using all sorts of assumptions about the heating system and its control that renders the output useless.
The Dutch system is better as is described in that paper. Am now looking for HR-107 rated boilers that really do modulate down and retain their efficiency (those tested under operating mode No 1) and will size everything such that cycle times are >10 minutes and return temps are <45C.
Still curious what a heating engineer would do.
*Assuming that the manufacturer is dishonest. pp92 of that document says that this is a reasonable assumption - all but one of the 70 boilers that the old DEFRA tested by 2007 were found to have grossly overstated their actual efficiencies whilst the so called independent certifying labs turned a blind eye. You also get lower total CO and CH4 emissions if you get the boiler stinking hot quickly then modulate down than if you warm it up slowly too.