All About Lightbulbs (not really)

[Less of a Draft as of 2013-06-24]
So my wife tells me that I know more than anybody should about lightbulbs. This is likely true, but I do find them interesting and useful to know about. I also like the thought of saving money on more efficient light bulbs. This quest for efficiency can be found in my previous post here. A few thoughts about light bulbs:

1. There are really four lightbulb options at the current time (though most folks are probably thinking only about three). They are LED, CFL, Incandescent and Halogen. The last is the one most people haven't seen in household applications (flashlights and cars, yes). Halogen turns out to be more efficient than incandescent and thus avoids the coming federal lightbulb crackdown.  They also have fair "normal" color temperature.
2. Some perspective, a 60 watt incandescent is equivalent to a 14 watt CFL and a 10 watt LED. That is a 6:1 energy advantage for LEDs over incandescent, but an only a 3:2 advantage over CFLs. See this chart.
3. The big difference between CFLs and LEDs is that the LEDs lifetimes are drastically longer. That is to say 50,000 hours versus 10,000 hours for CFLs (versus a measly 1,200 hours for incandescent).  This, I believe, is even a bigger deal in practice than is the case for the rated numbers.  See 5 below.
4. LEDs typically don't die (see exception at point 12). Their lifetime is considered over when their brightness is half of its initial brightness.  That said, I don't know what LED Lumen calculations are made on: beginning of life brightness or end of life?  (If you know - please write in the comments)
5. CFLs usually don't last as long as they are rated. This is very much dependent on the application, though. If you use a CFL in an area where the light is likely to get turned on and off frequently, the CFL won't last. Apparently, they need time to cool down before being turned back on again.
6. According to the DOE, the average house has 40 sockets in which to put lightbulbs. When I went through my house, I found a few more. To be specific, there are 186 bulbs in my house (not just sockets, but lamps too). That is a bunch.
7. If the average LED bulb costs $20, it would cost $3,720 to fully outfit my house in LEDs.  I am a fair way down that road, but I haven't gone all of the way.  Many of the bulbs just don't get used (or left on) enough to justify the switch.  I'm also positive that the LEDs  won't get reflected in the sale price of the house when I do sell it.
8. A bunch of bulbs in my house is an understatement. If all of the lightbulbs were on in my house and fitted with incandescents, there would be an electrical load of 13.5 kW (or about 112.5 amps).
9. The average household uses 2,100 kWh per year on lighting. That only equates to about $210 per year in electricity costs. Unfortunately, with all of our additional lights, I spend more than that per year just for the outdoor lighting (with the LEDs). Details are here.
10. The coming federal limits are actually already starting to take effect. See here. Best part is the labeling requirements which force manufacturers to provide some easily comparable numbers. The Nutrition Facts box on a box of cereal is a good analogy.  Unfortunately (IMHO), the requirements don't go far enough. They should provide a clear "cost of ownership" tag that shows how much it would cost to own the bulb over a 5 year period.
11. Don't install LEDs in your house if you think that you might move within the next couple of years. Unless, that is, you really really don't like to replace light bulbs. Or if you leave the lightbulb on all the time, like a night light.  Or, if you have turned into your father (maybe mother, but in my case it was exclusively my father) and have started complaining to family members for leaving lights on.  With LEDs, I find myself much less likely to complain.
12. I have used this (finding lights on in areas with no people) as a means of targeting bulbs in my house for replacement to LED.  Wandering around the house when I'm not normally there (middle of the work day) or in the morning or last thing at night is a great way to find the likely culprits.
13. Don't buy cheap LEDs.  I bought a set of three $7 GU10 (this is a reference to the connection type on the bottom of the bulb) at Amazon and the electronics gave out after only three months or so.
14. My go-to brand for LEDs is Feit.  I've never gotten a bad bulb from those guys and they have pretty cost-competitive products.  Costco carries a fair share of their bulbs and I was able to buy some from Home Depot, as well.  I am not paid by Feit, but I would totally sign up to be a "brand ambassador" or some such thing if it were offered to me.
15. LED light bulbs are still not available for every application that I have at home (at least not from remotely mainstream companies).  I have under counter lights that use bulbs typical denoted as "exit sign" bulbs.  They have a candelabra base and straight sides.  I can't use candelabra bulbs because there isn't sufficient clearance (those bow outward).  I will wait until I can buy either Phillips or Feit bulbs for these (burned too many times by off brands).
16. This may be obvious to most, but more efficient light bulbs save you money in two ways during the summer: (1) less electricity to generate the same amount of light (the super obvious way) and (2) less electricity to cool your house from the waste heat generated by your lightbulb.  You should have roughly 1/5 the waste heat to cool.  This isn't a huge downside in the winter when you are heating the house anyway, but heat generated by electricity is typically not the most cost-effective means of doing so.  I don't have good numbers on this, but this post suggests a high cost (as does my childhood where I lived in upstate NY and had baseboard electric and got to hear my dad complain all of the time about it).
17. Don't get hung up on the cooler color temperatures of a LED versus an incandescent.  LEDs have come a long way and you may very find you prefer the slightly cooler temperatures than the (IMHO) overly warm colors from old-school bulbs.
18. Most LED bulbs can be made to dim these days.  Doesn't seem like this should be a consideration for most usage, but they dim and dim well.
19. Beware of noisy LED bulbs.  I know folks who have found this to be a problem with a common buzz of fluorescents in the past, but it can be an LED problem, too.  This is particularly important for bulbs that will be right next to your bed or some other work space, probably less important for recessed lights.  I bought a Philips 12E26A60 LED bulb (reviewed here) and put it in the lamp next my bed.  I had to switch it out; it buzzed.  I swapped it out with a Feit 60W replacement bulb and am much happier with the silence (and its brighter).
20. While on the subject of Feit, I believe that their 40W and 60W bulbs are great replacements for your standard light bulb.  The only thing to be careful of with the 60W replacement bulb is that it is a fairly big bulb.  Make sure it fits in the desired space envelope.
21. Don't worry about the delay in an LED lightbulb turning on.  Many now have essentially no delay.  Others do, but when they turn on, they are all of the way on.  I've had many of CFLs that took way too long to actually brighten.  They were perhaps 30% bright instantly, but only 100% until about 5 minutes in (maybe it wasn't that long, but it sure seemed like it was).
22. Most LED light bulbs weigh substantially more than their incandescent cousins (with CFLs generally being in the middle).  Mostly, I don't think that you should have to consider this as a factor given that most light fixtures have between one and three lightbulbs.  That said, fixtures for which there are many bulbs (I have a chandelier that takes twelve 60 watt bulbs), this can add up to a substantial amount of additional weight.  Hopefully not enough to tear anything out of the ceiling, but consider it.  [TO DO: add actual weights of some bulbs]
23. Essentially all lighting fixtures have a max wattage on the bulbs driven by heat dissipation requirements (and the resulting fire hazard).  Using LEDs, you can turn a fixture that has a maximum 60 Watt bulb into something much brighter (and perhaps more functional).
24. As with all bulbs, pay attention to how light will be distributed from the bulb.  A "normal" light bulb is expected to disburse light in all directions, if not equally, close to it.  LED replacements can have somewhat different patterns and may not look good in your existing lighting fixtures.  For recessed lights, the LED bulbs should say if they are narrow or wide (120 degrees is about the widest you'll find).

Another Set of LED Bulbs

I finally found some GU10 LEDs that look like they should be useful.  Again, its Feit bulbs that seem to be hitting the sweet spot.  $11 per bulb, almost half the price of other brands (found at Home Depot in Reston, VA).  I bought three to start to replace the terrible bulbs I had bought at Amazon.  Read the reviews of that one - you won't buy it.  As for Feit, I'm pretty comfortable that they will last.

Below is my email addressing this subject (LED lightbulb replacements) sent to Tyler Cowen of Marginal Revolution fame.

Tyler,
Long-time follower of your blog.  I don't know if you respond to questions, but in the off chance that you do, I have a question that came up in my mind after I bought some LED lightbulbs at HomeDepot this morning.
I haven't seen this addressed elsewhere, though I will admit I haven't searched too extensively (but Google Scholar has been part of my attempt).  In any event, this might be an interesting topic for discussion given your ability to write clearly and distill issues down to their core.
So, consider the homeowner seeking to manage their lighting needs.  For instance, if a lightbulb goes out at the home, the owner would want to be able to replace it immediately and therefore requires a stock of unused bulbs.  However, on the last trip to the store, the homeowner decides to purchase LED bulbs due to a desire to (a) save money (b) not be the cranky dad that yells at the kids for leaving the lights on (from personal experience, this tendency is much reduced if the wattage is 1/5th that of incandescence) and (c) not change lightbulbs in the future.  (b) and (c) are hard to quantify, let's assume for the moment that their economic value is zero, along with the cost of changing a bulb.  When the homeowner returns home, she determines that all of her lightbulbs are functioning.  She has a nagging feeling though, that it makes sense for her to pre-emptively change bulbs to the LEDs given that her $20/bulb investment will otherwise go to waste.  What is the right framework for her to think about this problem (I suppose a standard DCF analysis could be used)?  Are there other investments that follow the same pattern?
It seems obvious to me that if what I said about (b) and (c) is true, she should immediately change the bulbs.  Relaxing the assumption that changing bulbs has a cost could potentially change that conclusion if the costs were high enough.  Do you agree?  Do you know if this has this been part of the energy efficiency discussion?
Regards,
Mark
http://stepbeyondgadgets.blogspot.com/

(Update 2013-06-15 17:15) Tyler was gracious enough to provide a quick response to my inquiry:

There is clearly some attention cost, and what economists call "S-s" rules seem to apply here, in any case I will ponder this more and blog if I come up with something...

Tyler

Looks like I have some research to do.  I will post later, if I think I have something interesting to say based on that.

Solar Insolation Monitor for Gardens

I just cleaned out a garden / pit of weeds and I am ready to put down some landscape fabric and mulch.  Pretty boring, yes, because I don't know what plants would do well in the sunlight / shade mix that this garden has.  This sparked an idea.

While others have come up with homemade solutions to measure solar insolation (basically, how much solar energy is hitting a certain spot), I propose an easy to use tool for landscape architects (and eventually homeowners, when they get cheap enough) to use for this purpose.  If this video or the The Garden Continuum's blog posting is any indication, there is not an existing tool that makes this easy for a would-be landscape architect.

Sunlite Tracker

The basic tool would look like a stake topped by a round half dome.  It would be a rough crystal ball, with small solar panels on each face.  The dome and the stake are separable with a quick twist of the dome.  The general notion is that you'd pop any number of the Sunlite Trackers in gardens around a residence to determine how much sunlight an area gets.  Consequently, one can determine with some accuracy what plants with thrive there.

In my conceived usage, the Sunlite Trackers would be set out for a week and then collected by the landscape architect.  The location of all of the domes would be recorded prior to their removal.  The Sunlite Tracker would record every minute, the amount of incoming solar insolation.  The architect would then pop a USB cable into the dome and download the information about the sunlight measured at the location.  It would be uploaded to the web-based service www.sunlite-tracker.com.  There, the data would be compared to sunlight measurements for the nearest weather station, so it can be normalized on a historical basis.  I think you do not want to rely on a week period that could have just been sunnier than normal to base your planting decisions on.

Then the website would display the set of trees/bushes/flowers that would thrive in those light conditions.  The architect could be confident that the plant recommended will be suitable for a given location.  If using enough of Sunlite Trackers, that confidence could be increased even in areas partially shaded by branches above.

Who should do this?

I think the business model could work like this:

• Landscape nursery signs up with Sunlite Tracker, enters all of their stock available.
• Landscape architect checks out some Sunlite Trackers and places in a customer's garden
• Landscape architect retrieves Sunlite Trackers and orders goods from the landscape nursery
• Landscape architect returns Sunlite Trackers to landscape nursery.

The key, then, is for the landscape nursery to see this as a means of expanding their sales and integrating more closely with their professional buyers - the landscape architects.  I think that I have a bit more to learn about the industry, though, before I declare this approach suitable.

One (or more) Step(s) Beyond:

Here are some ideas for extensions to the basic idea presented above (organized from least complex extension to most complex):

• The website should allow the architect to put in other filtering criteria, such as deer resistance and low maintenance.  This seems perhaps obvious, but many wholesalers have simple PDFs (some password protected to keep out retail customers like me).  I'm thinking something like the search page on Plant Select.
• Though a bit of a technological leap, one could make the stake smart and have it measure the acidity and moisture of the soil that it has been staked into.  While it doesn't seem like the acidity measurements need to be repeated over time, the moisture measurements probably should be.  And then the moisture measurements could also be normalized for the rainfall over the recent period.
• Also, set up the trackers to transmit back to the internet automatically with a mesh-based network of collecting and sharing information in the base tracker, all ultimately going to a Master-tracker that will be plugged into the home-owners wi-fi network.  This way, there is only the setup trip and the collection trip (with all of the processing and preparation done outside of the time spent at the home-owner's house).

Update 2013-07-02:

So it turns out that I'm probably just bad at doing research.  Folks have developed boards that will do this.  There are fairly well developed tools out there.  That said, I still don't think that they system has been put together in a way that would be easy and useful for professionals focused on the landscape design (rather than hobbyists or enthusiasts).

It looks as if what is needed most here is probably some calibration methodologies and software.  The "one-wire" setup appears easy to put together in the field and would have sufficient flexibility to cover a large or small garden setup.  The logging functionality exists, the computer interface, the ability to attach multiple sensors at the same time.  Its all there.

Though I'm sure it sounds terrible to a techie, good packaging around these products would make this equipment truly useful.

Fitbit fail

Those who have read my other fitbit related posts, know that I've had my fitbit since mid January 2013.  I wear it non-stop, even most nights to bed.  I'm odd enough that I feel like I'm not getting "credit" for physical activity if I'm not wearing it.  I should be fitbit's best & worst customer all at the same time.  I'm not super thrilled about my most recent fail with the fitbit.

OK, so its not the most important failure ever, but its still quite annoying.  The black plastic piece of my clip came off.  I'm pretty sure that this is a problem that should never happen.  We have engineers that can solve these sorts of things!

Nothing to do but order another one for $9.95 + $3.98 S&H.  Fun fun.

Update 2014-07-04:
The second cover that I had been using over the last year has finally given up the ghost.  While it did last longer than the first, the mode of failure was the same (the plastic piece at the end of the clip came off).  Annoying.  Here are the pictures of the old and new one - the design does seem to have been refined, though for me it mostly means its tougher to get the fitbit into the clip.

The rubber ripping and then peeling back has been an annoying issue for some time, but not a big enough deal to warrant replacement.  The plastic piece missing at the end is too painful to live with for long.

New one is on the right

Also annoying is that the price of a new one has gone up.  To $14.95 on Amazon (+ tax in the state that I live in).  I ordered a new one, but declined to get a new sleep wristband (that also needs to be replaced).  I found another product that looks like it will do a better job and cost less.  I'm considering buying this, instead: