IET logo
 
IET
Decrease font size
Increase font size
Topic Title: 18th DPC - part 8 - cable sizing
Topic Summary: please could someone check my maths and thinking?
Created On: 11 July 2017 10:29 AM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 11 July 2017 10:29 AM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

I've been trying to make some kind of sense of the proposed section on Energy Efficiency (and struggling a bit), and tried to work out the implications of some of the requirements.

For instance, 801.6.7.2 says about the Cross-sectional areas of conductors:
Increasing the cross-sectional area of conductors will reduce the power losses. This decision shall be made by assessing the savings within a time scale against the additional cost due to this over-sizing.
For cables, the chosen size shall be determined taking into account the cost of losses that will occur during the working life of the cable against the initial cost of the cable. A calculation method can be found in IEC 60287-3-2.

The I2Rt losses and limitations on future expansion of fed loads need to be considered for smaller conductors.

NOTE: In some applications (particularly industrial), the most economical cross-sectional area of conductor may be several
sizes larger than that required for thermal reasons.

I take this to mean we should not just pick the cheapest cable that satisfies CCC, v.d. etc, but take into account the cost of the losses (due to v.d.) too and pick the cheapest overall.

So I've tried some workings out - to keep the numbers simple, using a 10A circuit.

I think that losses and costs will be fairly proportional per metre of cable, so I can work things out on a per-metre basis rather than having the complication of a particular circuit length.

So my thinking goes like this, if I use 1.0mm2 cable, 10A circuit and presume a resistance of 16milliOhms/metre (yes I know it should be nearer 18 for 70-degree running but bare with me...)

Cable cost - say 20p/m, resistance 16 mOhms/m, so losses (I2R) = 1.6W/m so for 24 hours/day and 365 (and a quarter) days a year, that's about 14kWh, at say 15p/kWh that's about £2.10 a year per metre. For a cable lifetime of say 50 years that comes to a total cost of over £105 per metre per year. Have I really got that right?

Going up a size to 1.5mm2 (presuming a cable cost of 20p/mm2/m so 30p/m for 1.5mm2), I get a total (again over 50 years) of £70.43.

2.5mm2 gives £42.58.

4.0mm2 gives £27.10

6.0mm2 gives £18.73

10mm2 gives £12.52

16mm2 gives £9.77

25mm2 gives £9.21

35mm2 gives £10.01

So the cheapest overall size to use for my 10A circuit would be 25mm2.

(Obviously if the circuit wasn't drawing the fill 10A continuously the losses would be lower, so a smaller size could be justified - my spreadsheet suggests that if it ran 1 hour/day the "optimal" size goes down to about 4.0mm2.)

It seems to me that a "common sense" requirement is (if I've got my maths and thinking right) going to have some far from common sense results. Please could someone double check my thinking (or if you have access to IEC 60287-3-2 see if their method produces any more acceptable results).

- Andy.
 11 July 2017 10:34 AM
User is offline View Users Profile Print this message



dustydazzler

Posts: 1449
Joined: 19 January 2016

You lost me at 18th

Shudder and cringe
 11 July 2017 10:48 AM
User is offline View Users Profile Print this message



Alcomax

Posts: 248
Joined: 12 November 2009

Gulp...have not got near part 8 yet... maybe September, then too late!

Without checking maths, what immediately strikes me is that energy efficiency at consumers end with perhaps larger cable sizes than traditional approach is likely to be off set by greater amounts of copper and increased production costs. Does not appear very energy saving at planet level. Now I could be wholly off the mark with the next bit, but....what about DNO cable sizes? If those are not considered then what is the point?

Not really looked at the energy saving thing in 18th, but there appears to be a certain amount of Banana's about it.
 11 July 2017 12:18 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 8791
Joined: 23 April 2005

Andy


I have just done a quick spread sheet for a 10A load on 10m of singles with a cost of 15p per kWh for 8 hours a day for a year for various CSAs and using the cost of the cable from TLC.

My calculation says it would take 14.38 years to recover the cost of 10mm over using 1.5mm. In a real discounted cash flow you would need to put in an interest rate but I have done it a 0%.

Of course there is the added problem of getting 10mm singles in to your ceiling rose!

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 11 July 2017 12:22 PM
User is offline View Users Profile Print this message



leckie

Posts: 4395
Joined: 21 November 2008

 11 July 2017 12:54 PM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

For all you Amtech fanbois then read the cable analysis report - it tells you the cable power loss and the maximum power loss if fully loaded

As a designer we've been evaluating this for years - although from a cost rather than numbers of polar bears saved perspective

It's a useful tool to justify finger in the air cable sizing

OMS

-------------------------
Let the wind blow you, across a big floor.
 11 July 2017 01:11 PM
User is offline View Users Profile Print this message



broadgage

Posts: 2435
Joined: 07 August 2007

A little common sense may be applied in my view.
In most DOMESTIC jobs upsizing cables to reduce voltage drop and consequent energy waste is most unlikely to be worth while.

Very few domestic circuits have a long term load of more than a few amps, remember that the DNOs only allow on average 5 or at the most 10 amps per dwelling, and that this astonishingly low figure has a good record in practice.

There of course exceptions such as a relatively distant annex with a long hour electric heating load.
Another likely future exception would be a circuit for charging an EV since this may be fully loaded for many hours at a time.
Detailed calculation is a point pointless without knowing the charging requirements and the pattern of use.
I would suggest that good practice for an EV charging point could be to simply go "at least one size up" for the cable.
For example, 2.5mm would be the norm for a 16 amp circuit, I would suggest at least 4mm for EV charging.
A larger EV in regular use might cost £400 a year to charge, saving 3% of that by designing for 2% voltage drop rather than for 5% is worthwhile at £12 a year.

For large industrial installations it is often well worth upsizing cables to reduce losses.
For example a large submain might carry 250KW and saving 1% of that is worthwhile, 2.5KW is about 50 pence an hour, or several thousand pounds a year depending on the hours of use.
Installing perhaps a pair of 240mm SWA does not cost twice as much as a single cable.
 11 July 2017 01:20 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9568
Joined: 22 July 2004

I think the mathematical method is right - you have to make some assumption that electricity will rise in price more or less in line with materials over the years.
The 10A example coming out at 10mmsq or more is probably right, if 10mm cable is 10 times the cost of 1mm, and there are no other effects - it basically balances when losses over period cost the same as materials. There are obvious practical problems with this example as the cable wont fit.
But in our society to borrow from the future is cheap - which is why we may cut down many forests and don't plant new ones, as the return on the investment is great but is not taken up as it is 100 years away.

Cable loss is also dwarfed by other costs- even at max VD 19/20ths of the electricity is being used at the load, and your 10A 50 year load is costing hundreds of thousands of pounds - so why care about a few hundred either way - a 1% saving in on-time, or load efficiency is a much more cost effective way to reach the same or greater saving.

Fitting lights located so the light is spread around in way that you don't need so many watts of lighting may be a better design optimisation.

And then, in a house, is that heat wasted ? Not in winter, certainly, as it goes to reducing the fuel bill by not a lot. I agree if the house is cooled in summer it is a bad thing.

It is more complex than at first appears.

-------------------------
regards Mike
 11 July 2017 02:22 PM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

It's not that difficult to derive an annualised cost saving based on a specified payback period to determine the ECONOMIC upsizing required

Basically plot the cost of energy lost for the range of cable selections against the cost of those cable selections - from there you should be able to derive a "U" shaped curve - the nadir of which will tell you the most economic cable size to employ

Basically, the same technique as determining the economic cost of insulation for hot pipes for those looking for a suitable analogy

You can also determine a payback period - owner/occupiers will often have longer permitted payback periods than tenants would tolerate.

Contractors and spec developers would need some other reason to do this as they don't give a ****

I'm sitting at an industrial site at the moment with a 20MVA import capability owned outright - lots of fairly big constant loads like air compressors (6 x 375kW units on the 7 bar air system alone) and a lot more variable loads like welders (potentially 200 operatives all welding on a shift cycle) - I'm using a simple set of criteria to justify LV ring mains in 2 x1 x 4c x 400mm2 cabling that provides resilience, fault tolerance, expansion capability and a low(er) cost of ownership - capital cost is pretty far down the list in terms of any increase of the baseline cost - basically the client is willing to make some investment to save over a 15 year period (basically the predicted lifecycle of the current manufacturing plan)

The annual savings are pretty big - and hopefully the cost of cable can be reduced by purchasing a lot of the same size rather than many sizes

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 11 July 2017 07:25 PM
User is offline View Users Profile Print this message



sparkingchip

Posts: 10104
Joined: 18 January 2003

Just turn the lights off when they are not needed.

Andy B.
 15 July 2017 10:09 AM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

Thanks all - seems my suspicions aren't unfounded then.

Does anyone have any thoughts for a policy for reducing losses without cable sizes getting ridiculous? (and avoiding demanding customers take out a second mortgage to pay for the cable) - preferably keeping the calculations reasonably simple!
- Andy.
 15 July 2017 03:18 PM
User is offline View Users Profile Print this message



burn

Posts: 249
Joined: 06 June 2003

If I have understood it correctly, there's no need to force your customer to take out that second mortgage Andy.

We just have a duty to assess it and fully inform the customer of potential long term benefits. They can then chose to ignore the advice and take a short term cost saving.

It's a bit like having to have an EPC, you have be informed of the energy efficiency of a property but there's no compulsion to make any improvements.

Regards

burn
 17 July 2017 11:27 AM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

Originally posted by: AJJewsbury

Thanks all - seems my suspicions aren't unfounded then.

Does anyone have any thoughts for a policy for reducing losses without cable sizes getting ridiculous? (and avoiding demanding customers take out a second mortgage to pay for the cable) - preferably keeping the calculations reasonably simple!

- Andy.


Depends on the circumstances I guess - we have several clients, who include in their design guidance( usually mandatory) notes to the effect of the longer term economics of analysing I2R losses

It's reasonably easy to boil it down to

1 - Keep 20% spare capacity (creep)

2 - Operate at around 50C conductor temperature (typical trigger point for concern when thermal imaging)

3 - Sensible diversity assessment (ie build in a bit of margin)

Use that as a starting point and compare with the minimum criteria normally selected

Basically, I mentioned above the idea of "economic" up sizing of cables - once you set a reasonable payback period, you eliminate gross oversizing and typically end up with "a conservative" design as suggested above.

I usually try and target a few % saving of a big number of kWh/annum as being worthwhile (along with a whole load of other things to get demand down)

Have a rummage for the best practice guide regarding the economic sizing of pipe insulation - there is a simple arrangement in there that is equally comparable for cable sizing - ie energy cost saved, capital cost spent and draw the curve to get the nadir point based on specific payback (or annualised) criteria

It's the usual mantra of

1 - Get the demand down

2 - Meet it efficiently with minimal losses (in this case I2R losses)

do what is "reasonable"


Same idea as running two transformers in parallel at half load rather than duty/standby - you pay a bit more on switchgear fault level but loss is the square of the load so it's much more economic to do this rather than keep one on full load and the other hot but unloaded

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.

Edited: 17 July 2017 at 02:55 PM by OMS
 17 July 2017 11:55 AM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

You've got mail Andy

This covers the total losses in a cable (dielectric loss angle, sheath loss, circulating current loss etc

You can sensibly ignore these for a LV cable

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 18 July 2017 09:02 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

If I have understood it correctly, there's no need to force your customer to take out that second mortgage Andy.

We just have a duty to assess it and fully inform the customer of potential long term benefits. They can then chose to ignore the advice and take a short term cost saving.

I've read a few commentaries that suggest just that, but I'm having difficulty finding any words to that effect in the 18th itself - at least in the normative parts. The informative appendixes do seem to be based on that assumption - but still it's hard to see how the 'thou shalt' bits of part 8 can be justifiably ignored as it stands.

- Andy.
 19 July 2017 02:38 PM
User is offline View Users Profile Print this message



arg

Posts: 152
Joined: 18 March 2015

Originally posted by: AJJewsbury
Does anyone have any thoughts for a policy for reducing losses without cable sizes getting ridiculous? (and avoiding demanding customers take out a second mortgage to pay for the cable) - preferably keeping the calculations reasonably simple!


I think your original calculation was correct - except that you should be using the long-term average current rather than the rated current. So for a fixed load, scale it down by the fraction of the time you expect it to be turned on, typically a smaller factor than we use for diversity since diversity has to average out within an hour or so, while here we are talking about the average over the whole life of the installation.

For example, two 32A circuits, one for an electric shower and one for an EV chargepoint, the chargepoint needs the thicker cable as it's probably in use, say, 2 hours per day while the shower is perhaps 20 minutes per day - hence the calculations should be on the basis of average currents of 2.6A and 0.45A respectively.
 19 July 2017 04:10 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

I think your original calculation was correct - except that you should be using the long-term average current rather than the rated current.

Yes, I agree that the calculation should take into account the amount of time current is being drawn. I don't think that averaging the current is the way to do it though.

Losses are proportional to the square of the current (I²R) - not directly to the current itself. I think it's the long term average power loss you need to work on rather than the long term average current. E.g. say the 10A circuit was only on for 1 hour a day, the power loss would be 1/24th of what it would be if running continuously (say 10² x R x 1/24 = 4.16 x R Watts) . But calculating using 1/24th of the current continuously gives you a falsely lower power loss (say (10 x 1/24)² x R = 0.174 x R Watts).

I did do some calculations on average power loss basis - e.g. 10A for 1 hour per day - but that still comes out at about 4mm2.

For example, two 32A circuits, one for an electric shower and one for an EV chargepoint, the chargepoint needs the thicker cable as it's probably in use, say, 2 hours per day while the shower is perhaps 20 minutes per day


Yes, I agree entirely. My spreadsheet (for 50 years service life) suggests something like 10mm² for the shower and 25mm² for the charge point!

- Andy.
 19 July 2017 04:29 PM
User is offline View Users Profile Print this message



arg

Posts: 152
Joined: 18 March 2015

Originally posted by: AJJewsbury
Yes, I agree that the calculation should take into account the amount of time current is being drawn. I don't think that averaging the current is the way to do it though.


You are right, of course.

So apply the duty cycle factor to the cost saving after calculation rather than the current.

That also allows you to account for the oversized cable running at a lower temperature.

My spreadsheet (for 50 years service life) suggests something like 10mm² for the shower and 25mm² for the charge point!


I think 50 years is too long. Sure, the wiring might last 50 years under favourable circumstances, but many buildings are demolished before their theoretical life, and many others get re-purposed, refurbished, extended etc. causing individual circuits to have a shorter life in their original role. Perhaps 20 years would be more reasonable.

Edited: 19 July 2017 at 09:40 PM by arg
 19 July 2017 04:52 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

I think 50 years is too long. Sure, the wiring might last 50 years under favourable circumstances, but many buildings are demolished before their theoretical life, and many others get re-purposed, refurbished, extended etc. causing individual circuits to have a shorter life in their original role. Perhaps 20 years would be more reasonable.

Fair point - I picked 50 years just on the basis that PVC cables are meant to be good for at least 70 years and I'm still seeing rubber and imperial PVC cables that must be older than me. I'd certainly agree that some installation won't last anything like as long as that though - maybe 20 years is about right for a domestic kitchen between refurbs, probably a lot less for a branded retail unit (which seem to get a "refresh" ridiculously frequently) or rented office space where tenants change regularly. For the more unfashionable domestic circuit though 50 years is perhaps reasonable? I guess the thing here is to take into account the particular circumstances of the installation and circuit - rather than look for a-one-number-fits-all solution.

- Andy.
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2017 FuseTalk Inc. All rights reserved.

..