Replacing CFL bulbs with LEDs

I replaced the traditional incandescent bulbs in my flat some years ago and it was a no-brainer.  The compact florescent lights (CFLs) were coming rapidly down in price, were available in all the sizes I needed and of course consumed far less energy than the other heated filament bulbs they replaced.  I was happy to be saving energy and cutting my electricity bills.  All was good with the world, even though some of the bulbs took a while to "warm up" before they gave full output. I could live with that in most places.

Scroll forward to 2013 and I've now replaced a few of the first CFL bulbs with another new technology: mains power LED bulbs.  I need to replace some of the older CFLs that had failed over the 8 or 9 years I'd had them and Tesco (of all people) had new 4W LED bulbs on special offer at around £8 per bulb.  Now that's still a lot more than CFLs (which were being given away at one point and which you can quite happily get in pound shops these days).  However I wanted to try them out as I have four down lighters in the kitchen that are used quite a bit and where I could really do with the crisper, bluer, more immediate light from the LEDs.

I thus spent £32 on replacing 4 x 14W CFLs with 4 x 4W LEDs (compared to 4x100W reflectors that were there when I moved in over a decade ago!).  I liked the light and there's no "warming up" period (or at least none that I can detect).  Whilst the LED bulbs were more expensive the energy savings of having them on for a few hours per day on a regular basis means that I should be financially in front after 5 years or so, and the LED bulbs have a much longer predicted lifespan.

With that good experience under my belt, I started to wonder if there were other bulbs I should consider replacing in my flat.  The top contenders are the lounge standard lamp that is on for a few hours every evening, my bedside light (ditto) and the outside light above my front door (which could really do with a brighter light with no warm up whilst I'm trying to go up or down my iron staircase). As a result of discussions at the Footpaths Group at Loughborough University I started to wonder what the embodied energy of the various bulbs is and that's where things got alot more complicated.

Embodied energy is the energy required to actually make and distribute the product. The more complex the processing required to make a product and/or the more raw materials requiring high energy processing, the higher the embodied energy is.  The US Government funded an analysis of the embodied energy in different types of bulbs:

• Incandescent bulbs require an average of 42MJ per 20 million lumen hours,
• CFLs require an average of 170MJ per 20 million lumen hours,
• Current LEDs require an average of 343MJ per 20 million lumen hours.

The "MJ per 20 million lumen hours" might seem to be a rather odd set of units, but it is basically used to standardise the amount of energy (MJ - mega-joules) over a fixed light output (the "per 20 million lumen hours") irrespective of the actual light output of the resulting bulbs.  This lets you compare the embodied energy required to replace, for example, a single very bright incandescent bulb with several CFLs or LEDs with lower light outputs (in lumens).

Now this looks bad for LEDs but the same report also shows the energy used actually lighting the bulbs. The traditional bulb consumes 15,100 MJ per 20 million lumen hours, the CFLs use up 3780 MJ per 20 million lumen hours and current LEDs sip just 3540 MJ per 20 million lumen hours.  This confirms that replacing traditional incandescent bulbs with either CFLs or LEDs is a good thing to do.  Even though both CFLs and LEDs have higher embodied energy, they consume about a fifth of the energy required to provide the standardised light output, and the energy used in lighting the incandescent bulbs dwarfs the embodied energy.

The tricky call is replacing CFLs with current LEDs though.  Whilst the energy used in powering the LEDs is 240MJ less per 20 million lumen hours, they require 173MJ more to make.  So they do win out in the end energy wise but the break even point comes rather late in their life.

I'm probably still going to replace my high use CFLs with LEDs.  However waiting a while may help: mains powered LEDs are still a relatively young technology and the energy use, embodied energy and prices are all likely to fall over the next few years.  My existing CFLs are still pretty good, so its probably best to wait until they break, and then make the switch to LEDs.

Shedding some light on commercial lighting

After my last blog post looking at electricity usage in Loughborough as the first part of an Energy Descent plan for Loughborough, my chum Martin asked what proportion of energy is wasted lighting up offices and shops when they are closed?  This is a very good question as I've often bemoaned the lights left on in shops when I walk through the town centre late at night.

At the University where I work some of the offices have motion sensors attached to the lighting, so the lights switch off 15-20 minutes after the last people have left (or if I sit too still in the evening!).  Not all the University buildings have this feature, and even in buildings where some offices do have it other areas don't (for example I often have to turn off the lights in the kitchens and mezzanine service areas on my way out, as they aren't on sensors and most people don't seem to know what the little switch on the wall does, despite having energy saving advice stickers attached to them recently!).

So how much energy are these shops and offices using, and can we work out (or at least guesstimate!) what sort of energy is being wasted by leaving the lights on all night?  The first question is quite easy to answer, because the DECC have a handy summary of "service sector"  energy use(which includes retail and commercial offices, as well as sports facilities, government offices, health, education and few other bits and bobs.  It doesn't include the "industrial" users such as factories, mines and construction).

The summary tells use that lighting is responsible for 21% of the sector's total energy use, which in turn was 18,357 thousand tonnes of oil equivalent in 2010.

Now what we need to do is convert "thousand tonnes of oil equivalent" into something a bit more familiar - good old kWh.  The conversion factor is quite simple:

1 ktoe = 11630000 kWh

So the 18,357ktoe is equivalent to 213491910000kWh, or  213.49191TWh.  The lighting is 21% of that, so 44.8333011TWh.

Now that's a fair chunk of energy in lighting all those offices, shops, schools, warehouses, etc.  But what proportion of that is unnecessary?  I guess we now need to think about what we mean by "unnecessary"?  There's some obvious ones:

• Lighting the internal areas of shops and offices when there's nobody working there,
• Lighting external areas when there is nobody around in the wee hours,
• Uplighting buildings for purely aesthetic reasons (some of which seems to be left on during daylight!).

But what about the less obvious wasting of energy for lighting?  For example there's some evidence that since the 1950s there has been an increasing amount of "over illumination", especially in parts of the retail sector.  If we've got people using more light than they really need for tasks, then that's wasting energy.

There's also the plea that some people raise that lights are left on at night for security.  The idea is that you'd be able to see criminals doing their nefarious deeds if the rooms are all illuminated.  However it doesn't seem to work out like that - if there are few people around in the middle of the night there aren't going to be many witnesses.  Even if there were passers by, seeing lights being turned on or torches flashing in an otherwise normally dark building may be just as good a give away of naughtiness happening.  Indeed the criminals usually need light to do whatever it is they shouldn't be doing, so leaving lots of lights on can actually help them.  Councils that have dimmed or turned off street lights have sometimes recorded falls in levels of crime for example.

So how do we find out how much of that ~45TWh of lighting energy is being burnt pointlessly?  That's not easy to answer, especially in our increasingly 24 hour world where some shops and offices never close.  I've had a good old trawl through the Interwebs with Mr Google, and there's a distinct lack of hard data we can use to determine the answer to this question.

One paper I did come across had an interesting graph for a real London office that showed that the base load (ie the power used all the time, irrespective of whether people were using the building or not) was about 60% of the peak load when the workers were in. I've seen similar base load charts from the University's Sustainability team's energy monitoring of campus buildings.

Now this base load power use wasn't just lighting but also other electricity usage such as leaving IT equipment on and air conditioning running 24/7.  Even so, it shows that there's a lot of overnight energy being used in typical offices and we already know from above that office lighting is one of the highest electricity users overall.  If the 60% of the peak lighting is begin left on overnight across the country and not being used between (say) 10pm and 6am in most buildings, then we can guesstimate the wasted power.

Lets call the amount of base load lighting power used each hour p.  Then we'll use 8p at night and 16p of base load during the day.  To keep things simple lets assume that everyone comes in at 6am and turns on all the extra peak stuff at once, which then stays on until 10pm (it doesn't - this over estimates peak loading somewhat so will make our base load figures a bit low.  But heck, this is very rough guesstimate territory now!).  This means we'll use an extra 16 hours of the 40% peak power, which is 16 times 4p/6.  Lets assume this happens every day of the year, and use this to work out out a value for p from our known total lighting demand:

45 = 365*(8p + 16p + (16 * 4 / 6)p)

p = 0.003556375TWh = 3.556375GWh

So this guesstimate tells us that every hour we've got a base load lighting usage nationwide in service sector buildings of about 3.6GWh.  Doing this 8 hours a day, 365 days a year would mean a consumption of around 10.3TWh each year.  That's a big number - about a quarter of total lighting energy use!

So how do we get wasted energy use down? Unfortunately lighting isn't high on the list of priorities for some companies, despite them ending up spending hundreds, thousands or even millions of pounds on the energy it consumes.  Their accountants look at the capital costs of changing lighting systems and if the pay back period is more than a couple of years, it gets passed over. Heck, some retailers even have a problem with closing the doors in the Winter!

Another problem is that many service sector companies are tenants, and often energy use is lumped into a standard service charge.  The building managers.owners have no impetus to reduce energy use as the service charge covers their expenditure, whilst the tenants may be unaware of what fraction of the ever increasing service charge is due to energy use and what is due to other constantly increasing costs such water supply, sewerage or insurance.  This is not only an issue for how we encourage businesses to become more energy efficient, but also whether we can persuade them to invest in micro-generation technologies as well - tenants often can't and owners don't see the need to.

Of course as energy prices continue to rise, the pay back periods for installing smarter lighting systems and profits to be made from reducing or turning off extraneous lights will improve, so hopefully we'll have more companies taking notice of the lights the have on in the middle of the night.  And we can carry on nagging organisations that we work for or shop at to cut down on the obvious energy wastes such as pointless lighting.  After all we all paying for it somehow.