Friday, February 27, 2009

Energy Harvesting - not ready for prime time?

Nice brief summary article relating to energy harvesting technologies:

http://www.greensupplyline.com/214301888

I've got mixed feelings about his assertion that energy harvesting isn't ready for "prime time" yet, though. While I will agree that it's not close to being a technology that fits every application yet, there are certainly some nice niches that fit well.

For example, there are a lot of buildings that can be retrofitted with energy saving lighting controls (California's Title 24 regulation in particular) but the installation cost is high in order to route wiring for occupancy sensors in the room (or even worse, installing a sensor on every fixture). In this case, a wireless, battery-free, energy harvesting motion detector (other manufacturers with the same concept: 1, 2) is a perfect fit. While the equipment cost is higher than that of a wired system, the installation cost is much lower, possibly less expensive overall, but certainly competitive. The sensor harvests solar energy, stores it (enough to run for two days or so in darkness) and operates the occupancy sensor and the radio transmitter.

In contrast, and ultrasonic sensor uses thousands of times more energy, so to make an energy harvesting vacancy sensor (or occupancy sensor) would require a very large solar panel, enough to make it impractical. So much energy is required, in fact, that it would make a lot of sense to turn the sensor off when the lights are off (unless automatic on is required -- CA Title 24 doesn't allow for that) or at least reduce it's operating duty cycle.

On a related note, there is a movement to revive magnetic coupling to charge cell phones, laptops, and the like. Inductive power coupling is nothing new, it's been around for years. I have a waterproof cordless phone that charges without any electrical contacts, I understand that many electric toothbrushes recharge the same way. In much the same way that Apple made touchscreens new and exciting with the iPod Touch and the iPhone, companies are going to revive magnetic coupling so you can put your laptop on a special pad on (or embedded in) a table or desk and charge without a cord.

While it's convenient, the relatively low efficiency of this power transfer method runs counter to some of the goals of the green movement. On the upside, however, the stray fields from these devices may provide a new mechanism for energy harvesters. Adding a small magnetic pickup coil to an energy harvesting device may, in the future, provide sufficient energy to run these sensors without the need for a solar cell, mechanical, or thermal harvester.

Monday, February 9, 2009

LED vs. CFL vs. Incandescent

This article I read on edn.com, while it contains nothing I haven't read elsewhere, is a nice summary of the whole migration toward LED lighting. I hope to see LED lights that approach the efficacy (efficiency in terms of lumens per watt) of a good fluorescent light (around 100 lumens per watt). One named in the article was getting closer (about 80) and I've heard of laboratory tests of prototypes that are in the neighborhood of 150 (yes!).

I still want to see an intelligent way to retrofit existing dimming bulbs with dimmable CFL and LED fixtures. I think a bidirectional communication protocol between the bulb and the dimmer might be a way to do it, perhaps over the existing power line. The problem with that method is the extra cost, complexity, and reliability of adding that hardware to the bulbs and the dimmers. A dimmer would probably need to detect the connected load, and either use traditional dimming (for regular bulbs) or leave the power fully on and communicate the dim level, letting the ballast at the load control the brightness.

Perhaps the communication can occur by doing phase cutting, much like regular dimming, but use a 95% brightness level for one digital level (zero or one) and 100% brightness for the other value, and use, in effect, a serial data stream over the power line.

Long term, however, low or medium voltage DC power to the lighting loads, with digital control at the point of load, would be more efficient and have fewer points of failure than the current methods.

Just a few random thoughts. I hope to come up with a more coherent post in the near future. It's been a very busy few weeks, so I haven't been able to write a more organized, thoughful post than this.