Bulbs, power, lumen, and savings how do you know if coned?

[ericmark]

This is nothing new, around 1992 I was working on the build of Sizewell power station and my boss decided to use fluorescent tubes to light the tunnels, all were 110 volt units at ratting of 58W so quick calculation just over 1 amp for two tubes, so supply was 16A so 32 lamps would be just over the limit, so decided to fit 25 lights to each 16A supply.

Lights fitted 2 hours latter a call they were not working, armed with clamp-on ammeter I went to find out why. Answer they were drawing nearly 25 amp not 16 amp. At this point I took a lamp apart to have a look at it.

There were 5 components in the lamp, an auto transformer, a ballast, a starter, a capacitor, and of course the tube. The auto transformer lifted the voltage from 110 volt to 240 volt and the rest of the lamp was a standard 240 volt fitting. The auto transformer was not centre tapped but marked 110 volt and 127 volt, by moving the tapping from 110 and instead using 127 the current used dropped a lot but last 5 lamps would some times not fire up, so 20 lamps were wired 127 volt and last 5 at 110 volt and the system then worked well.

At the same time I experimented removing the capacitor and noted the power used increased quite a lot, but lamp still worked with no noticeable difference. This means a set of lights could use well over the rated wattage if the capacitor fails. It also showed how important voltage is. Today we use switch mode ballasts with 110 volt fluorescent lamps which auto corrects for small voltage variations. However be it fluorescent or LED without some voltage regulation they can use a lot more of less power than shown on the packet.

The LED lamp has an advantage over the fluorescent as if the voltage or current is lower than what they are designed for they will still work. Although like the fluorescent over voltage can raise the current used out of proportion to voltage rise. So one should always go for under voltage rather than over voltage.

The LED like the fluorescent can be supplied with a switch mode power supply which will auto correct the voltage. This is what you will likely find with a LED tube to replace a fluorescent tube. However with small lamps the cost of the switched mode unit starts to get silly, so they use a much simpler method to reduce the current, either a capacitor with an AC unit or a resistor with a DC unit.

This gives us another problem, with the capacitor the frequency is important, I have seen 12 volt LED bulbs marked 50 Hz. But having worked in Hong Kong I know their voltage is often 250 volt, so when making an LED bulb it would make sense to design it for 250 volt rather than 230 volt.

Another problem in the UK is we wire lamp to lamp and the switch does not have a neutral taken to it, this can cause a small voltage to be supplied to the lamp even when off, the cure is to leak away some of the power to stop fluorescent lamps flashing or LED lamps to come on dim.

So most white raw LED's will produce 100 lumens per watt, but built into a lamp this can drop to 60 lumen per watt. The smaller the lamp the higher the percentage of power thrown away to stop lamps coming on dim when switched off. So light a room with 10 x 3W lamps and it will waste more power than a single 30W lamp. However size matters, if you have 10 lamps you are emitting light from around one square foot of surface, but with a single bulb then likely down to a few square inches, so although the ratted lumen output of the 10 x 3W is lower than the single 30W lamp the room is brighter with the 10 x 3W lamps.

I have photography as a hobby, so I use the camera in doors a lot and can see the exposure required. Not measuring lumen, but it is measuring the amount of light. So a 16 foot x 10 foot room with camera set at 100 ASA (ISO on newer cameras) with an aperture of f2 the speed of shutter is around 1/15th of a second when using 10 x 3W LED bulbs. However there are glass bowls around each bulb which defuse the lights and all bulbs are facing up so light reflected from ceiling.

However I am sure it is possible to use a camera to compare the output from two different bulbs. I was reading the post By Big-all about his set of bulbs, and the comments on how he had been conned getting bulbs which were under spec. When I first got CFL I did measure the power, and an 11W lamp did use 11W, but that was before I had an energy meter to be able to measure power factor or power used over a set time.

Most of my bulbs today are SES (E14) in the ceiling lamps, but in table lamps I am using B22d bulbs, and to use an energy meter I need to be able to plug it in. Also both LED and fluorescent lamps get dimmer with age, so a new bulb will likely always show up in a better light (pun intended) than a old one.

But maybe we can devise a method to compare? Maybe not with a fluorescent tube, but with a bulb could we not design a standard test to compare bulbs? A picture with a lamp at set distance away with reflector behind lamp and the light meter reading on the picture for example.

Could we build up a method to test a bulb. Even following the manufacturers data we see bulbs rated from 60 to 100 lumens per watt so likely in real terms it's even more than that. Where I see the problem is I can't really correct voltage, some where I did have a variable transformer, but even if I found it, and boxed it so it can be used, unlikely anyone else is going to do the same. So best we can really hope for is noting voltage on the energy meter.

So is it worth doing, or are there simply too many variables for the data collected to be any use.

What does the team think?

[kellys_eye]

Pointless.

All that most people are interested in is the 'energy saving' so the comparison is already 'standardised' by virtue of the claimed wattage each device reports. Only the geeks amongst us make reference to lumens or ft lamberts or whatever to associate the overall efficiency based on light spread/output.

Unless the industry cares to introduce such a standard (as if ) then devising our own isn't worth the trouble.

Must say, though, that don't agree with your adjustment of settings for the fluorescents in that tunnel - it should have been properly designed to use standard fittings at standard voltages etc to simplify replacement and eliminate problems caused by swap-outs being 'non-standard'.

[AAA.Handy.Man]

ericmark

Many thanks for that 'Article'. Whilst some of it went over my head, other parts prompted me, by the wonders of the InterWeb, to follow up 'terms' which are of everyday use to many, but new to me.

I also particularly liked your analysis:

So most white raw LED's will produce 100 lumens per watt, but built into a lamp this can drop to 60 lumen per watt. The smaller the lamp the higher the percentage of power thrown away to stop lamps coming on dim when switched off. So light a room with 10 x 3W lamps and it will waste more power than a single 30W lamp. However size matters, if you have 10 lamps you are emitting light from around one square foot of surface, but with a single bulb then likely down to a few square inches, so although the ratted lumen output of the 10 x 3W is lower than the single 30W lamp the room is brighter with the 10 x 3W lamps.

Many thanks for an enjoyable, and challenging, read.

[aeromech3]

Interesting article eric; I had avoided LED in UK due unit cost and what I perceived as a dim alternative, due the heat generated by them was not as wasted as some would say, but as saving on central heating use. In my new Thailand build, I went for CFL 11W rating 550lumen daylight 'bulbs' as they were reasonable at £1.60 and was I quite content with 5 in an 18sqmtr ceiling, however the local Tesco Lotus now have 6W LED 'bulbs', and at £1.35 each I gave them a try and they are most definately brighter, yet only 470lumen and the quoted saving of 8W against equivalent CFL (14W) and 34W against GLS (40W standard incandescent), of course a long life is rumoured, as well as the cooler operation which is a Big benifit in this part of the World.

[big-all]

as you will probably know i use a flow meter i paid about £12 for mine
http://www.toolstation.com/shop/p90072? ... 20Meter%20
i plug everything in and go through several cycles and make my own ratings lables measuring off/standby/record/playback/complete cycle etc
so i know what everything cost and what price you can put on an action or lack off action
for example my front room telly was a samsung but seldom used the main tv was an identical sized sony i measured the sony it used about 100w in use where as the samsung was only 44w so now watch the samsung saving about £30 a year in running costs as its on for about 12hrs a day give or take
i could further save a further £35 a year by watching programs on a laptop monitor on the desk beside my laptop as the closer picture is the visually the same size but choose not to as the volume is great at the laptop but not enough when i go into the kitchen or bathroom where i can still see the telly with all doors open
i also leave the tv plugged in on standby as at 1.1w consumption for the other 12ish hours would save less than £1[about 80p]

i also worked out a load off washing including tumble drying cost between 25-30p a load so saves a load off time by never hanging stuff out to dry
now the machine is 1600 spin speed and the dryer senses the moisture in the washing so very very efficient

i measure all my led light bulbs and write the actual consumption and date off purchase on the bulb
i go by which bulb gives the level off light i am after but the lowest consumption for example the bulb above my head is rated at 12w and actually uses 5.8w and would rate the output at around 75-80w equivalent compared to the previous toolstation bulb rated at 12w used 12.6w and perhaps 90-100w equivalent and wee bit too bright
now with the right information you can make lifestyle choices and decide if turning off all light normally in daily use on and off as you use them is worth the effort or worry but decided as my total light bill will be between around £12-£18 a year dependent on turning on and off as used so not worth the effort involved to save £5 a year
all my calculations are based on 1w for 10 hrs a day at around 14p per ah/unit off power will cost 50p a year
1x10=10
10x365=3.650 units =50p at 13.7p a unit
by making the calculation simple with lots off assumptions it allows you to easily work out the cost benefits within quite a close value with little effort

[AAA.Handy.Man]

I am mightily impressed....

[big-all]

I am mightily impressed....

to be honest we all spend time on areas we find fulfilling and interesting
we wish to make it easy to understand and useful in general
if some find it useful it makes it worthy off the effort

[ericmark]

I have also used the plug in energy meter, one thing it does which clamp on ammeter does not is record time. I decided to measure what the fridge, fridge/freezer and freezer used, but mine and rest of family.

It seems they have a problem with damaged insulation, should it get damaged then water in the atmosphere will slowly form ice which conducts better than the insulation so the motors run far longer than they should, in general the motor runs for around 1/2 to 2/3rd of the time. Manufacturers have to publish running costs, so a little maths will tell you how much it should use in 24 hours, when the insulation goes this figure can become far higher.

So I measure and find a very small 32 litre freezer my mother has using well over the rated value, in fact if I take the run current and time it by 24 that was the power used in 24 hours, in other words it was not switching off.

OK identified there is a fault, however would get same result if thermostat stuck, so step one was measure the temperature, however non of my thermometers go down to -18C so need to buy one, I tried about every food shop in the area, nothing. Then found -18C is freezing point of brine so boiled up some water mixed as much salt as it would absorb then into freezer, and it went solid, so seems thermostat US.

Next tried to find some one to repair it, they wanted to visit, would not give me an idea of cost, but minimum charge was £70 and freezer new cost £100.

So two things, one I sent for a STC-1000 thermostat for my beer brewing, and two sent for new thermostat for freezer, the STC-1000 arrived first, so temporary set that up to control freezer. Then thermostat arrived and used STC-1000 to check temperature, good job I did, found at mid position temperature just -8C anyway all set up and running again.

After this started doing some maths, how much energy wasted with a faulty freezer, only 65W running motor correct ran for around 16 hours per day under fault 24 hours, so 8 hours at 65W so 520 watt hours per day, that's 190 kWh per year, at 16p per kWh that's £30 a year wasted. So if I had the repair done for me it would have paid for its self in 2.5 years, if I had bought a new freezer then it would have taken 3.5 years to break even.

Hang on though the freezer was only just out of it's two year warrantee, so if the replacement lasted same as original then it would not pay for its self. OK there is more than money involved, ready meals being too cold means likely not hot enough after heating and too warm and they may go off. But a thermometer would be a better tool than an energy meter, it's the temperature that's important not the energy used.

OK I also used it to work out size of bulb required to keep my beer at the right temperature, by keeping heater to minimum it means less hysteresis. However in the main the energy meter is a toy, it does not really help, OK the example of swapping two TV's yes it works, but it's rare for that to happen.

The problem with most energy using devices is they interact with other things, be it light bulb giving out inferred heat, or the washing machine costing less to run. My washer at home has inverter drive and water jets and should clean the cloths with less energy used to my mothers washing machine. But how can you measure it? You would need a set pile of cloths to wash and test the washer using the cheapest cycle needed to clean them, so likely an average over a year, but use a hot wash when a cold one would do, or use special soap when a cheap one would do, or have to do a wash twice because cycle selected was not long enough, and every thing is up the creak.

In the real world we wash longer than required and hotter than required simply as we don't know what is required and have to put cloths through a second time is the last thing we want.

[big-all]

to be honest the savings we are talking about both you ericmark and me are pretty much micromanaging savings with a small return on outlay but we enjoy the challenge enjoy getting worthwhile and not so worthwhile results
but at least we know where the savings are and can help others withought them wasting there lives finding out or spending money on a monitor
far greater savings as you said can be had from wearing another layer turning heating down by one degree or better draught reduction