Once you apply enough fiddle factors, it becomes no better than reading the rating on the supply fuse and multiplying by 230.

I can see load shedding becoming more important in the future.  My new EV charger came with a current clamp for the meter tails.  If I start using too much power inside the house, the EV charger will tell the car to slow down its charging.

Interesting, I've only given it a quick once-over, but if initial reactions are of any use to you...

1. "Traditional methods, such as those outlined by the IET, often result in inaccuracies" - couldn't agree more. The forum has had loads of questions over the years where someone has diligently applied the OSG final-circuit based approach and come out with ridiculously high numbers (>>150A for a simple domestic). Often the suggested remedy is to point out that it's appliances that are loads, not circuits, so the OSG method works best if you ignore what circuits are actually installed and instead imagine the minimal circuit arrangement that could supply the actual loads (e.g. one ring final circuit per 100m² of floor space, rather than one per (maybe 40m²) storey plus another for the kitchen), probably a single lighting circuit, small loads on FCUs off the socket circuit, not on their own circuits from the CU, etc. Oh and LED lighting at a sensible wattage (e.g. closer to 10W instead of the traditional 100W per lampholder).

2. "focusing instead on the actual number of occupants" - that looks like it'll pose difficulties to me, since the actual number of occupants can vary significantly during the lifetime of an installation. In the last 10 years we've gone from 3, to 2 to 4 and back to 2 (plus a dog). I wouldn't want to be in the position of having to re-wire (or even re-check/modify) the installation or re-negotiate a DNO supply every time we added someone to the household.

3. Per final circuit.. as above (1.) - that seems to be introducing an unnecessary degree of randomness. Many people prefer extra circuits for all sort of reasons that don't imply extra loads (often minimising inconvenience in the case of a fault, following the natural layout of the building (e.g. per floor circuits), or just personal preference (e.g. radials rather than rings - so maybe 2off 20A circuits rather than 1off 32A)). I've seen houses go from a 4- or 6-way fusebox to a 10 or 15 way CU during a re-wire - with no actual change of load at all. Often a more abstract metric (e.g. floor area) can provide a simpler general use load approximation, although maybe non-linear (larger homes may tend to use less per m² than smaller ones).

4. circuit demand factors ... Security Alarm 1 - combi boiler 1 - door bell 0.2 - those feel incredibly inaccurate - a security alarm probably takes <10W - a combi <200W, and a door bell really is pretty negligible, if they're fed from 6A MCBs to reckon them at over a kW (or 276W for the door bell) and still get the right overall answer at the end suggests to me that they might be hiding a mis-calculation somewhere else.

5. "This overestimation is a significant concern when engaging with DNO networks and applying for grid connections ...  Outdated calculations tie up valuable grid capacity, causing consumers and electrical distributors to pay for unnecessary upgrades." I'm not entirely following the logic here. AFAIK the DNOs have their own methods of calculating loads and diversity (ADMD) based on their own experience, which traditionally come out with figures getting towards order of magnitude less than OSG calculations, e.g. somewhere in the region of 8-10A per household (for non-electric space heating), but it does depends on a quite a number of factors, including number of customers on each supply.  I thought they only take the numbers declared on the application form as a signal to look out for anything unusual (e.g. large disturbing loads), they certainly don't size their mains and transformers by just adding up all the declared MDs.

  - Andy.

Interesting ideas indeed - I too agree that the OSG methods of sizing for MD are way off reality for any normal domestic.  Personally I find building size more useful -100A fuses feeding half a dozen 1 bed flats each with 63A fuses seem to work OK for ever.

Can I suggest a 'maximum likely occupancy' - using the no of bedrooms or something (allowing for multiple occupation of double beds I suppose.) to sidestep the uncertainty about the no of folk in the house and their loads over the life of the building.

One needs to be careful with how the results of such a calculation may be used - the maximum demand, as used to size a transformer feeding a street of 50 houses per phase and a row of small shops with 3 phases, can be a lot lower that the no. of houses times the ADMD of each house, as the peaks in load never coincide, and if they ever did the thermal time constant of the transformer to reach a dangerous temperature is a few hours.

However, the meter tails in each house should really at least use the ADMD or maybe the MD for that building, as they heat up in a few tens of minutes.

Mike.

PS it's a detail but the spelling of 'revision' in the footnotes, needs, shall we say, revision - such things stick out and make you wonder what other detail has perhaps been missed.

Hi Mike,

Yes, good catch with the spelling, unfortunately one of my weak points in life, the document is still in its infancy, and good feed back like that only helps it become stronger. Yes I do check items etc thing get through, I'm not posing to be perfect part of the process in making the document enhanced is sharing like this and review good productive feed back like yours.

The new calculations are trying to address more is the  installation MD when were looking at network planning (so when we put a application in to the DNO for a new EV charger, ASHP, etc we want to reduce the number of limitations on load or upgrade charges to the customer when they are not required). Were not looking at promoting these to be used as final circuit MD calculations which will have an impact on conductor CSA and OCPD. 

I tried the building size in an earlier version of the formula, but that variant didn't prove to be useful in it accuracy to take forward.

For the meter tails, going from my comment above, the CSA of the tails is already standard normally for domestic installs, but the CSA of the tails would also be subject to the up stream OCPD rating.

Don't get hung up on the spelling, other folk can fix that. (me perhaps ;-)

My main point was that 'network planning' does mean different things, transformers for the street see an averaged load, but the branches that feed just one or two properties are likely to see a more peaky load.
Keep up the thinking..

Mike.

Hi Andy,

thanks for the feedback, see below responses.

1. Yes agree, the two modes of calculation are trying to address two things 1st being accuracy of max MD for the installation, and 2nd simplicity in the calculation method which included the reduced number of data points.

2. So, agree but the calculations are for MD for the installation at the point of connection for the DNO, not the installations final circuits. these new calculations are needed for when your applying for a EV charger/ASHP etc to be connected and have more realistic/accurate property MD, rather than dictating current carrying capacity design of final circuits.

3.  Yes, can see your point and one myself would have originally agreed with until I started to see the data and reflection in the calculations (on the initial trials, I was even able to challenge a data sample as the calculations loads were way out of accuracy and asked a revisit for the electrician to ensure we hadn't missed any other DB's, are sure enough there was one hidden in the loft!) so yes we can have loads of DBs etc and loads not changing but this will also be relucted in the demand factors applied. (when we increase the test installation site to 200 plus maybe as the review grows we will see other trends immerge)

4. again, were assessing MD at the installation point of connection, not final circuits specific current profiles.

5. So, giving the positions I have held and worked in I have seen first hand the effects of overestimated maximum demands and the effects of such, as final financial burden to the customer who just wants an EV charger or the DNO, needlessly upgrading infrastructure where it isn't required. The ADMD, isn't that accurate either as the review shows, we are doing the calculations as per the AMDM in the scrutiny comparative reviews with other know MD calculation assessments and methodologies.

Hi Simon,

who would use the fuse rating and multiply by 230 ? 60A fuse (60x230=1,380) ? are you saying to use this as MD?

can you elaborate more on the fiddle factors? the formulas are being proved against real world data logger  load current profile recordings for MD, so any formula variations have been rationalised and justified based on factual field trial data. which the to modes are out performing the IET and National grids MDAD calculation modes.

Yes, load shedding is good, but load shedding tech cant be used for customers who want ASHP and over predicting MD does lead to upgrade cost for the installation, these calculations based on their accuracy levels, show that in a lot of cases the MD is a lot lower than originally assumed and can remove the financial burden of supply upgrades and take away the choke on a lot of schemes and projects.

(Pedantic note: that should be 60x230-13,800)

My response was somewhat tongue-in-cheek. 

I'm assuming that you already have the supply, have already ordered the supply, or are looking at a price list for a new supply and picking the highest you can afford.

Many years ago, I heard of a "constant" called Finagle's Finagling Factor.  Oddly, I can't find any reference to it on Google.  Suppose you're doing a calculation, and you know what answer you're expecting.  But the answer you get doesn't match the one you expected.  It's clear you need to apply a correction factor.

If the answer you want is X, and you got the answer Y, then the correction factor must be X/Y.  Because that way, the answer comes out as Y * X/Y = X.  Problem solved.

We know that an average sized house with non-electric heating will run happily on a 60A fuse.  If your load estimate comes to 150A, then it's probably wrong.  So keep coming up with "correction factors" to apply to your estimate until you converge on Finagle's Finagling Factor.  In this example, it's 0.4.

That's probably not going to work so well for an ASHP on a 60A supply.  But given that heat pumps are all computerised these days, I don't see why you can't tell one to turn off for a few minutes while somebody has a shower.

Hi Simon,

okay I now get your view point no issues.

so the calculation formulas are fixed, in that we don't keep tweaking to get a favourable result.

from our live logging sessions the results have shown MD to be a lot lower, even with and EV charger, air source heat pump, 4 plus person living etc the MD isn't as high as you assume, and the two modes of new calculation do a great job in a true reflection of installation MD at the point of connection to the DNO.

My aim for these two calculation methods is to give industry a solution which is not only quicker and simpler to perform but also significantly more accurate, and this will be then back up by real world data (Empirical Validation/Modelling Process), by monitoring property's under test for two weeks with a Fluke power quality analyser (the master data sets and backing tables to this report the raw data will be made available to institutions like the IET or academia for review and scrutiny them self's).

I do like your idea around load shedding being added to the ASHP, think were long away from such power management systems being utilised within property's, but also as we are proving with factual data in a lot of cases the MD is lower than previously assumed so there isn't a need to restrict the property's usage at given time, this wont be the case for all but a lot of stand household will truly benefit from this.

I have seen this firm hand at a country level scale of how many people wanting to go green but they cant or flocking out of pocket needlessly because some electrician used an out dated mode of MD calculation!

the estimation method works best for electricians who have first hand knowledge of the number of site DB's and the specific final circuits info. the occupancy method works best for DNO's and network planners