ECN Forum
A customer of mine has a large warehouse lit with metal halide high bays. We are considering replacing them with fluorescent high bays.

Since the fluorescent high bays use an electronic ballast, could we have a problem with harmonics ?

Suggestions ? Theories ?
I have installed T-5 fixtures (5 -4ft bulbs per fixture) in place of 400 watt metal halide high bay fixtures with great results. The lighting is slightly brighter, and seems to be of much better quality.

I have not made any measurements, but I can't imagine the electronics for the fluorescents to be any worse than the electronics used by the HID ballasts.
I've replaced several 400W MH fixtures with 6 lamp T5's in several different businesses. Each installation to T5's improved the light (more) and resulted in less electricity (aprox 1/3 less) usage. No one has complained of any harmonics or causes any electrical problems with other adjacent tennats. Defineatly the way to go, especially if your changing out ballasts on the MH fixtures, plus you have advantage of instant on. Our price was about $215 per fixture plus 6 blubs at 6.50 each. Highly recommended.
Ballasts=feedback harmonics. Not sure the dif between the two types on your application. They have meters to measure such things, but they are around $2K. Might be able to rent one if you're lucky though. Good luck!
Change them ASAP. The lastest and greatest fixtures do not present Harmonic problems.
Light output will be much better. Also each individual fixture can be ordered with a PID sensor for even better savings.
Our electrical engineer only designs warehouses and factories with MH hi-bays only if specifically requested... It's been T5HO's (mostly 3 lamps per fixtuer) for a little while now and no complaints.
We use a lot of the Highbay 4, 6 and 8 Lamp T5HO Fluorescent Fixtures - with Reflectors and Parabolic Diffusers.

They are a much better choice for wide spread area lighting than an HID Highbay or Lowbay, as the linear Lamps have a good light spread - especially with the reflectors and Parabolic Diffusers!

The THD (Harmonics) of the Electronic Ballasts does not result in anymore L-N Load Current then what is produced by a typical CWA Reactor based HID Ballast.
Maybe an additional 5% if the Circuits are loaded to Maximum, and the Ballasts are "Noisy" (>20% THD).

The typical tradeoff for a Lighting Retrofit (old Probe Start Metal Halide fixtures replaced with T5 HO Fluorescents) is a drammatic reduction in drawn Kilowatt Hours, while still retaining the same level of Light (and a much "cleaner" light output too!).
Clients receive conservation + reduction benefits, and typically the retrofit pays for it's self within 2 years.

The Ballasts driving the Lamps on the fixtures I typically Specify / Design with, are "Programmed Start" instead of just plain old "Instant Start".
The Lamps are started in the "Hot Cathode" mode, and as they are run, the Cathode Heating is reduced accordingly.
This results in excellent Lamp efficiency + effecacy (read: high, stable light output), best life operation, and since the Cathodes' heat is reduced during operation, the total drawn wattage is reduced at that time

Standard "Rapid Start" methods are great for Lamp operation, but power is wasted by continuously heating the Cathodes at the same level.

"Instant Start" methods use "Cold Cathode" operation, but the Lamps are not running at their optimum output.
The Cathodes are not heated, so there is no "Parasitic Load", or "wasted power" to heat the Cathodes; the tradeoff is lower power drawn through the Lamps, and results are lower Light output.

Scott
Scott made mention of T-5's having a 'cleaner' light. I know, that's not an easily defined term ... but it's still true, regardless.

Comparing 400 watt MH to 5-tube T-5's, in actual use the overall level of lighting seemed identical. That is, the side of the shop with the T-5's did not seem any brighter than the side next to it, with the MH. Yet ... on the T-5 side ... the light seemed 'softer,' colors clearer, shadows lesser. You just seemed to be able to see things better - especially when doing detail work on a bench.

The power savings are very real, as well. The same light, or better, for about 1/3 of the electric bill. Other advantages were a lesser load on the air conditioning (HID gets HOT!), and the fact that the fixture never goes completely dark. That's what happens when the only bulb or ballast in a fixture goes bad; the T-5's I saw had 5 bulbs and 2 ballasts.
Adding to Reno's last post;

The term "Cleaner Light" is just my interpretation - not any justifiable term.
The Light output appears more uniform and less distorted.
This must be contributed by the 20kHz Lamp operating Frequency, the Parabolic Diffusing apparatus, the use of "HO" (High Output) Lamps, and the 5,000K Color Temperature Lamps.

The power savings is the most apparent thing between the two technologies.
Along with this, there are other benefits such as what Reno mentioned - Lamp failure does not take out an entire fixture, only one or two Lamps of an array will be affected.

Additionally, the total heat produced is somewhat less with Fluorescents, and the Fluorescents have an "almost instant On" characteristic, one which is not in comparison with HIDs - even PSMH Electronic Ballastry - Auto Dimming.

The usage of Linear Fluorescent Bay Fixtures -vs- an HID Bay Fixture depends mainly on the Environment, as follows:

*A An HID Fixture may be better suited where tasks are in specific points - so the light is concentrated more in the task area, rather than a wide distribution.

*B An HO Linear Fluorescent Fixture may be better suited in Warehouses (like Racking Aisles), because a wide distribution of light is needed overall, instead of specific point intensities.

Scott
Good stuff guys!,
I am also a big fan of the newer T-5 fluorescent fittings.
I've used them in quite a few applications, including factories, the "yards" of a slaughter-house, they even make a street-light-type fitting that uses these lamps.

The only down side to these fittings that I have experienced, is temperature changes can cause the light output of the tubes to vary slightly, but it is not a big variation.

I would be interested to see how a M-H lamp actually stacks up against a T-5 fitting, using real-world analysis, not the information given in lighting company literature, where the tests are often done in some sort of a laboratory.
I believe the first place to start would be comparing the output of the two lamps in lumens/watt.
Any takers?
Mike (Trumpy):

Quote
The only down side to these fittings that I have experienced, is temperature changes can cause the light output of the tubes to vary slightly, but it is not a big variation.


This is a very valid statement, as when we use Fluorescent Lamps inside of Walk-In Freezers, where Temperatures are -10ºC / 14ºF and down, the light output is reduced by 10% to 20% - depending on the ambient.

The Ballasts are Program Start, and incorporate an additional heat generating circuit, to keep the Ballastry + Lamps somewhat warm, without becoming a large contributor to heat.

On the Walk-In Freezers, I typically will set a "Target" illumination (in Foot-Candles) for 120% the desired average Foot-Candle level.
For example, if the desired level in the Freezer is to be 30 FC, with highly reflective walls, and fixtures mounted in rows at +20 feet A.F.F., I will set my application's target average level at 40 FC.

If the reflectance is less (darker points, like what would result from brown cardboard boxes on shelving), and/or the mounting height is >30 feet, then I adjust the average FC target by as much as 160%
For example, if 30 FC is the desired target average in the Freezer, I set my application's target to 50 FC average.

No Manufacturer has photometrics for the expected reduction in sub zero environments.
They all say their fixtures work in those environments - which they do, just no one has taken the effort to give actual derated levels.

We have been kind of finding out as we go, to come up with the +20% - +60% figures. So far they have been rather accurate when applied to the typical fixtures we specify and install.

Scott
Scott,

Dealing with "cleaner light," sometimes a simple term like this helps others understand the more technical term. How many understand the proper term "Color Renditioning Index" or the more cryptic "CRI." This relates to the visual problems that you mention and what I tend to call harsh light because it hurts my eyes after a time.

CRI - is an approximation based on the complete light spectrum of daylight and that is because we have no official definition of full spectrum light (the sun puts off allot of light that we can not see and even window treatments change the color of sunlight). There is also allot of flawed research into this area and the benefits of full spectrum light effects. Suffice it to say that a definition of CRI=100 was given to the standard incandescent bulb and is the maximum possible value.

Typically lighting designs should not be less than a CRI of 70, and I personally like to light offices at a minimum CRI of 80. Once above 90, the lumens per watt output of the available sources drops off and should be reserved for special cases such as museums or locations where color is extremely important.

Now looking at standard light sources and typical CRI's:

Metal halide: 65-70

High Pressure Sodium (HPS): 20-25

Ceramic Metal Halide:80-92

Highend HPS: 60-85

Fluorescent: 70-95 Some as bad as 50.

Compact Fluorescent(T5): Typically greater than 82, but I had the misfortune of finding some that I bet were about 50 for my house good thing they had an average life of maybe 300 hours (what a waste of money).

LED: Varies to much, look at each individually. New technologies have impooved so much that I am not qualified to guess what they are reaching now for some white sources.

Mercury Vapor: 40-50 (some as low as 20)

Thus, with metal halide fixtures at CRI=65-70 versus T5's at 82 plus the perception is allot more pleasing to the end user.
Re the metal halide retrofit, has anyone out there had any experience with induction lighting? I have heard figures of 100,000 hour lamp life, and one third the consumption of metal halide systems, and apparently they are well suited for retrofits, but this could just be sales talk.
Originally Posted by sabrown

... we have no official definition of full spectrum light...


Actually we do--the black body radiation of a black body of the specified temperature. The sun is pretty close to a black body radiator. Incandescent filaments are good approximations to black body radiators, which is why they have a CRI of 100%.
© ECN Electrical Forums