[OT} Electricity usage

In message , News writes

This morning, for the last couple of hours it has been fluctuating around 500W.

A couple of laptops on, a PC for most of that time. a server (about 40w idling) , various wallwarts for router, VDSL modem, wifi AP, a couple of network switches and the normal other stuff scattered about. Tropical fishtank, A fridge and small and large freezer and probably a couple of lights on somewhere.

We also have a similar size elec consumption, so independently of this thread I have been having another go at looking at where ours has been going. We do have a lot of occupation of the house though. 2 adults, 2 kids, but we home educate, so often 3 of us a round in the daytime as well.

Keep in mind this is a lighting circuit (or more to the point a pair of them), and all the lights on the circuit were off at the time of testing. I have also eliminated the PIR and the dimmers from the suspect list. I will check the charging current on some of the emergency lights next...

In message , alan_m writes

Looks useful. Thank you.

In message , Dave Liquorice writes

Brilliant. My monitor does exactly that - or at least it would, if I could get the poxy PC to see the poxy monitor ...

Meanwhile, having turned off almost everything, I got down to 84w usage. A couple of phones and the router.

Turning stuff back on was interesting. Difficult to be exact because I don't know when things like freezers are coming on or off, but :

I really need to do this with everyone else out of the house. What is slightly worrying is that it seems to be obvious that this usage is us specific rather than building specific. In other words, if we moved from this Victorian pile to a modern, insulated 2 bed bungalow and took our current habits with us, electricity consumption would not change.

P.S. Good to know that this thread is making others check their usage and habits :-)

If those are the tubes that you said are on all the time, there's 5 units a day gone. Just 40W, for 24 hours, is a unit.

Indeed. Two are on one switch, and three on another. Starting last night, the three were off! These are tubes in the kitchen, and they really do need to be on during the day. The kitchen is quite large, but poorly lit. Time to think about alternative lighting, I think.

Continuous or just when charging? Room for improvement here.

Are these on a lot? Look to making changes here.

Is one of them old? Or both, possibly? May be worth re-investing.

This is OK.

Agreed :-)

Except changing the habits or lights is perhaps easier than moving.

In message , Tim Streater writes

Normally they seem to be only when charging-certainly that's what ours do.

We've a fairly old about (20 years?) under counter freezer that draws

Sounds like and old Pentium 4 :-)

Yeah, unless it's a gaming rig, no need for it to anything like that. I think the main PC here draws something like 100 -120W max?

Well, it certainly did that with a couple of £8.99 digital "Power" Meters my dad bought in Aldi about a decade back. The experience of "The 'cheap digital watt meter" as demonstrated by these two examples rather soured my view of the more expensive 'cheap' digital wattmeters that were available at the time such as the classic "Kill-A-Watt"(tm) and its European variants in popular use by the home energy brigade (a group largely dominated by the Yanks).

This cheap Aldi meter (model #DEM1379 for the more curious) seemed to work more like a simple VA meter, especially for loads less than 100W varying by up to +/- 10W on some 10 to 20 watt electronic loads (as measured with a real wattmeter - the MetraWatt model I've mentioned elsewhere).

In short, it was fekin' useless for checking the stuff I wanted to check (wallwarts, PCs and similar type loads). Even treating its readings as VA measurements didn't seem to account for its unpredictable behaviour.

I guess they'd rather skimped on the quality of the ADC chips and the processor that was supposed to calculate each product of voltage and current sample taken at a fast enough sampling rate to sum the results to generate a true(ish) rms value of the power consumed by the load.

It just disturbs me a little to see mention of a +/- 10W accuracy for a current(?) model of digital watt meter. That suggests it might simply be a revamped DEM1379 (which, btw, also used a couple of AG13 button cells to provide backup power to maintain the units consumed log).

You can test with a table lamp using various CFLs or LED and Tungsten filament lamps to check its behaviour at the lower power levels with a mix of purely resistive and reactive loads. If the meter is better than the +/-10W accuracy implied, you can trust it to give meaningful readings when checking the power consumption of your IT kit otherwise, restrict it to toasters and kettles and buy a better one to measure the "Tricky Loads".

From personal experience, the only models of digital watt meter I'd care to recommend are the L61AQ[1] and the N67FU models sold by Maplin for a penny or two shy of 20 and 10 quid respectively. The N67FU, unlike the L61AQ (a Europeanised Kill-A-Watt) can show sub watt figures. Above the 1W mark, they both display to a tenth of a watt. The accuracy is about the same for both[2] so I'd recommend the cheaper N67FU as the 'Best Value' buy.

Presumably chargeing their attached laptops, not alot you can do about that. Modern switched mode power supplys/chargers are pretty effcient.

With thse devices it's not so much the amount they draw, the laws pf physics dictate that to shift X amount of heat from A to B you need Y amount of input power. How often the heat pump needs to run to keep the inside cold is what matters, and old relatively poorly insulated device will run more often than a modern AAAdouble plus highly insulated one.

Or a 1 GHz single core Athlon that I'm using, though I think it's only about 150W. It probably does chomp through 2 to 3 units/day though. One day I'll change it, *IF* I get a 120 W saving (ie 180 W down to 60 W) that would be £80/year, payback in 3 years ish. Hmmm...

Agreed, swap out the banks of 2 and 3 2' tubes with a single 5' for each. 2 x 58 = 116 W...

words, if

Yeah but the point is household energy consumption is really down to the life style of the occupants rather than anything to do with the building, provided you exclude space heating.

If all the lights that are currently on (16) were 60 W incandescants they would double our current load to 2 kW. Most of those 16 are CFLs so I'd say the actual lighting load is less than 1/4 of the 1 kW we are taking.

OK, if you insist ;)

All valid considerations. I agree these sort of fixed connected loads are problematical with regard to verifying their actual consumption. Digital controllers usually only take a watt or two, the CH pump usually somewhere in the region of 50 to 100 watts (the old Grundfoss UPS 15-60 3 speed pump I pulled from my CH system a few years ago is rated at 45, 70 and 105 watts for the three speed options - can't recall what the replacement is rated at). The Honeywell motorised diverter valve takes 6 watts and I'd expect less than a watt for a motor driven time switch.

The most practical way to get such readings, without intercepting the feed to connect a watt meter, is to shut everything else off and monitor the suppliers KWH meter for a sufficiently long sampling period (counting disk revs or led blinks) to at least keep the error down to less than 10%. In the case of the CH, this might take half an hour or so (circa 100 watt load, mainly the pump).

You might be able to make one up but the main problem will be accessing the neutral connection (although you could get away with using a handy earth bonded point as a substitute neutral - the meter draws only microamps from the neutral for the voltage sensor). It'll be tricky enough making up a fuse cartridge 'plug' to divert the live via the meter's current sensor but doable with a little ingenuity.

That may work ok if the voltage probe relies on a galvanic connection to the live (or has that as an option. Some of these clamp on current sensors designed to clamp around meter tails also use the clamp itself to sense the live voltage via capacitive coupling which may be compromised when clamped round a thinner cable (reduced coupling of the electric field (voltage) which will give a lower reading than actual.

It's unlikely to be more than a watt or two when 'switched off'. If you really need to 'know', you can always fumble around to gain access to the 13A socket.

That's certainly a workable solution. It just needs a modicum of care and common sense. Presumably the main aim here is to log the consumption over a few days of use (you know what the lamp takes, you just don't know how much run time it'll typically be clocking up - the rest of the circuit standby consumption is unlikely to be more than a watt or two, but you'll be able to verify this to remove such 'mystery' from your estimates)

When I installed the security light a couple of decades ago, I picked up the power from the top floor ring main via a 5A fused connector unit. I just didn't fancy having a 500W halogen switch on surge risking a blown lighting fuse on the upstairs and basement lighting circuit (or, for that matter, the ground floor lighting circuit).

If I wanted to connect an 'energy meter' I'd have to use a cheap extension lead as a 'breakout box for the plug in meter and wire it into the FU box for the duration.

Some of the older ones could be using as much as 4 or 5 watts just keeping themselves at nice and toasty 40 to 50 degrees. A good quality bell transformer could well be wasting as little as half a watt or less with the lamp in the illuminated bell push consuming a further half watt or so with the bell itself perhaps taking no more than another watt or two on each press of the bell push.

My solution for the front door bell was to use a battery powered ex-GPO trembler bell (with a wireless bell push wired, via blocking diode across the bell push contacts so the 3 x 9v PP3 battery pack bought in poundland didn't backfeed into the 12v powered wireless sender unit used to signal the repeater 'bell unit' I keep in my upstairs office - the GPO trembler bell is mounted right onto the front door frame).

I think I replaced the decade or so old 16 by AA cell 24v battery pack with the PP3 pack just over a year ago. I certainly expect the poundland '27 volt' battery to last at least 12 months or more, making the running cost far cheaper than a typical mains powered bell (even the most efficient of mains powered illuminated bell push doorbells is going to struggle to compete with that level of economy :-)

Again, if you want to check the actual consumption of the mains powered bell, you're going to have to intercept the mains feed to plug in your 'energy monitor'. for the stakes involved (around a fiver's worth on the annual electricity bill), I'd be inclined to leave this exercise on the back burner for now.

Getting a handle on your PIR controlled flood lighting, however, would, imo, be worth the hassle involved since it is potentially a higher stakes game (a 3 to 5 hundred watt lamp running for 3 to 10 minutes at a time, maybe a few dozen times a night can add a significant amount of running cost comparable to your daily use of the electric kettle!).

Nope..

Nope, thats on a spur off the ring main..

Well if you pull that fuse it drops to Zero so some current must be going somewhere down that particular circuit.

I'll stick me bench Fluke meter across the fuse holder when a get a moment as see what that sez...

cheers..

Thinking about this, I think my last attempt might have been before quite a lot of rewiring was done, and now there's probably more separate circuits than before, so using the supplier's meter and selective turning off of some circuits might actually be easier...

And, I can also work out how much power one unit on a circuit uses (even if it's hard to get at) if I can use the plug-in thing to monitor some other units on the same circuit (eg temporarily running them off an extension lead connected to the plug-in thing).

So eg I could use the plug-in meter to measure the fridge freezer's use, but make sure only the oven was also on the same ring main, turn off everything except that ring main, then oven use = supplier's meter - plug-in meter.

Good point...

But for example for the alarm's spur, there's a 13A socket right next to it; I could take a neutral wire from a 13A plug (with nothing connected to its L or E pins).

Maybe I should take the faceplate off the fused spur unit and interecept the supply much as I suggested for the PIR outside lights.

These things designed to go around a L meter tail; are they influenced by an adjacent (very adjacent in T&E) neutral?

Low wattage - I certainly hope so, but I remember a certain amount of desperation when trying to work out what else was on in the house, to account for more load..

Yes, I'd want to measure several nights' use.

Before the rewise, the bell transformer was screwed into a brick subwall well under the floor - so (in line with my other thread, unlikely to be much of a fire risk!) - and just wired into the nearest 13A plug - so easy to measure... if only I had done then. But during the rewire the sparks mounted it on the wooden board that the CU is on, making its connection nearly impossible to get at...

It's always irritated me that illuminated bell pushes have bulbs that burn out reletively quickly, spare bulbs are/were hard to get and fiddly to change over. When I came here the cable through the wall into the back of the bell push was friable and really short... I put a surface mount box and blank plate on the inside wall near it, into which the supply from the transformer & the cable out to the bell sounder & the bell-push wires now all come... so that I could change the through-wall cable easily AND have some spare cable coiled up in the box so that there was a few inches of spare to work with each time I took the bell push off the wall.

Thanks for your thoughts!

Well they're not that bad as long as THEY TURN THE THINGS OFF afterwards*

M.

• After say cooking my dinner, so that's a good trade-off.

In message , News writes

Attempting to prove a negative is not positive fault finding. I would turnoff the power during the day then go round to the shop and see if they are running around like headless chickens wondering why their tills have stopped working and their freezers are all defrosting.

In message , charles writes

Or even better 0.23

Notionally or in reality?