Thd and pf for led retrofits

[Cletis]

I have a customer highly concerned about
An led retrofit raising his thd. I know this is now non-linear and sinusidal. I would think lower power consumption down 65% and higher pf led drivers would offset this??? We are lowering facility 100k w/hr


Sent from my iPhone using Tapatalk

 

[Electric-Light]

I have a customer highly concerned about
An led retrofit raising his thd. I know this is now non-linear and sinusidal. I would think lower power consumption down 65% and higher pf led drivers would offset this??? We are lowering facility 100k w/hr

Sent from my iPhone using Tapatalk

Lowering kW usage together with adversely affecting lighting performance is easy to do and its a popular trick of many retrofit sales outfits.

His concerns are valid.
A good lighting ballast:
presents to the line side THD <10%, PF > 0.98

Do not pass down flicker inducing 120Hz ripple current to the lamp. If you can play with camera settings and cause "banding" in image, flicker is getting through. As a reference, the flicker present in high quality external CFL and T8 ballast is comparable to studio lighting. Most consumer LED bulbs fail this test. Many LED lamps are worse than magnetic ballasted fluorescent as they allow flicker to get injected into light output in order to meet power factor requirements while cutting corners to shave cost on ballast BOM cost.

Many power bricks for 12v/24v LED lighting, especially those from non lighting brands are consumer electronics grade and power factor on those can be 0.5 to 0.6 with a THD of 80-120%.

There are high end (not "contractor select") extremely expensive LED retrofits sporting three figure lm/W, good endurance(edge that lasts rather than just stay sharp for a short while) and those parameters.

 

[Dennis Alwon]

It is nice that you guys know these terms but it would be good when introducing fairly new term abbreviations that you explain , at least, what they stand for.

 

[Electric-Light]

It is nice that you guys know these terms but it would be good when introducing fairly new term abbreviations that you explain , at least, what they stand for.

Power factor is the ratio between the watt and multiplying volts and amps. In the past it was mostly caused by phase shift which is easily correctd with capacitors. In modern days, it is also caused by irregular current waveform dawn by electronic devices such as computers and LED ballasts and it is not easy to remove.

Industry standards quickly clamped down on commercial fluorescent lighting and kept up the power factor and kept down the allowable THD. Now fluorescent lighting has a PF >0.95 and THD often less than 10%. Given relatively large power supplies used for computers, they're steadily catching up for units used in corporate/education environment.

Regular CFLs, fluorescent ballasts marked household use only, "ballast bypass" type LEDs and cheap low voltage AC adapters for LEDs are still commonly available in gross polluters (THD 80-120%, PF 0.6). The minimalist transistorized LED ballast built into "ballast bypass" type retrofit frequently omits the costly active PFC and struggles to satisfy power quality and flicker free output at the same time.

Total Harmonic Distortion is one of the power quality metrics.

What to ask when choosing LEDhuge missing pieces: flicker percent and flicker index and FCC conformance.

 

[Dennis Alwon]

It was the THD I had not seen-- total harmonic distortion-- Thanks

 

[electrofelon]

I have a customer highly concerned about
An led retrofit raising his thd.

What is he worried about? I am not aware of any POCOs that charge for thd/non linear loads. Besides, these things are pretty good these days. Why would an electronic fuorescent ballast be substantially different?

 

[Electric-Light]

What is he worried about? I am not aware of any POCOs that charge for thd/non linear loads. Besides, these things are pretty good these days. Why would an electronic fuorescent ballast be substantially different?

What's the point of LED change out again? After grasping a good technical understanding and theory of supposed savings, did an accountant suggest it was a wise idea to spend/finance a large sum of money for this particular vs keeping, or vs another alternative? He's talking about a lighting system where 100kW is only a portion of lighting load so this is probably an industrial facility.

Distorted waveform also have a poor factor. If the plant already has high harmonic loads that are essential to their business, you can't blame him for not wanting lighting to add more harmonics. Transformers and wiring losses are based on VA rather than wattage. If it's non linear, the loss is actually even higher. Draw enough non linear current and voltage starts to become distorted. Distorted voltage causes motors to lose efficiency.

This is a good website to look at if you want to look at a bunch of excuses Lunera Resources

Excuses for having poor PF caused by mismatched load

Glossing over inability to provide maintain full output in the installed environment due to fragility

HID lamps have a high lumen density. A proper retrofit can only be called such when it can provide intended performance in the environment it is meant to be used. A grill can work two ways. Run at full power, and cycle on and off with to satisfy the thermostat or throttle the wattage continuously to hold the set temperature. A lighting product should only go into thermostatic limit (limp mode) during abnormal operations to prevent damage.

due to LEDs being prone to heat damage, some LED products have become dependent on regularly tripping LED fry out preventive mechanism while being advertised with specs that it can not maintain in all the intended advertised applications. LED retrofit bulb suspended by the base in a bell jar struggles to provide output matching HID while rejecting heat adequately to avoid cooking thermally delicate LEDs and its ballast. "Thermal IQ" "intelligent thermal management" and such are marketing gloss over to creatively say their in-situ retrofit do not have enough durability to work at full performance and it places them in limp mode rather than burning out. It's the only alternative to cycling on and off in the interest of avoiding fried LEDs. Other options exist such as phase change cooling but it can not be done without bloating the price tag more.

Computers often make use of variable speed fan to keep the CPU cooled under different load. Thermal fold back takes over only when increasing fan can not maintain the temperature limit and it usually only happens due to abnormal room temperature or blocked airflow. If it happened all the time, that computer is not of the performance purported in the specs.

Dedicated LED fixtures can operate without regularly activating LED burnout protection and it does so by giving LEDs low thermal resistance path to the fixture body and LED burnout protection should only activate under unusual conditions, such as heater vent blowing onto it.
Some are rated at L90 (hours to 10% LED decay) rather than L70 (hours to 30% LED decay), but these tend to be highly expensive.

LED ballast/LED driver occasionally tries to get by without utilizing cost adding active PFC but at the expense of impaired light quality in the form of producing excessive line frequency Flicker.

Insist on
*power factor better than 0.95
*THD under 15%

Flicker:
*not to exceed 20% flicker percentage
*and not to exceed 0.10 flicker index
*Thermal throttling should not activate under intended use during normal operation.

It's not a significant difference from power quality perspective but insisting on greater than 0.95 PF & under 15% THD power quality performance rather than settling for >0.9 <20% tends to weed out performance compromise, cost cutting ballasts.

The compromise is letting huge amounts of line frequency flicker bleed into the light output which causes motion blurring, saw blades that appear to not be in motion, and interfere with cameras. Fortunately requiring 0.95 PF and <15% THD makes it more challenge to fit through the loophole.

Here's an example of compromise LED driver ballast that often finds use as disposable ballast for screw in LED bulbs.

http://www.onsemi.com/pub_link/Collateral/AND9094-D.PDF
"LED Ripple Current < ±35%"

"For high powered applications like fluorescent light
ballasts, It is common to use a dedicated boost front end
converter stage to deliver power factor greater than 0.99.
This is much higher performance than required in an LED
bulb or down-light and adds an additional power conversion
stage."

Well they forgot to explain 35% ripple causes huge amounts of flicker and the so called "much higher performance than required" design does not degrade light quality by infusing flicker. Not having excessive flicker barred meant invitation for LED products with more flicker than magnetic ballast fluorescent lights.

 

[electrofelon]

Boy, i am sorry you have had so much trouble with leds: fragility, shutdown from overtemp, burnout, flicker, drivers with horrible thd......I havnt had a single one of these problems and all my customers have been super happy.

 

[iwire]

Boy, i am sorry you have had so much trouble with leds: fragility, shutdown from overtemp, burnout, flicker, drivers with horrible thd......I havnt had a single one of these problems and all my customers have been super happy.

It is odd.

At this point the company I work for has installed tens of thousands (yes that many) LED fixtures of all kinds and we are just not seeing the failures.

Oh for sure there are failures but no more so than the failures we were seeing with electronic ballasts for fluorescents.

I don't know if Electric-Lights wife ran off with an LED salesman but he sure seems to have a vendetta against LEDs to the point of ignoring any success stories .

 

[Cletis]

Isnt there some kind of basic harmonics Meter that would read this for under $500? Ill just start taking readings before/after
Prova-19 ? $289 maybe


Sent from my iPhone using Tapatalk

 

[peter d]

I don't know if Electric-Lights wife ran off with an LED salesman but he sure seems to have a vendetta against LEDs to the point of ignoring any success stories .

At the very least, it's amusing to watch him frantically rearrange deck chairs on the Titanic. :lol:

 

[Electric-Light]

Isnt there some kind of basic harmonics Meter that would read this for under $500? Ill just start taking readings before/after
Prova-19 ? $289 maybe

You can rent a proper harmonics analyzer for that price. A cheapest unit that provides uncertain accuracy is fine for knowledge but not if the reading has to support spec non compliance. It's better to layout strict specification requirements prior to installation and tie it to completion and payment terms rather than waiting until completion to test and bicker.

Those considering LEDs could simply copy and paste this to the LED solutions sales rep

power factor shall be 90% or better, THD shall be under 20%.
flicker performance shall be <20% flicker percentage and <0.10 flicker index.
Failure shall not exhibit flashing or strobing during or beyohd warranty.
Thermal foldback shall not activate during normal operation. Normal means conditions customary to the location installed. Any LED decay compensation that intentionally increases wattage with accumulated usage shall be disclosed clearly.

About 0.98 power factor with under 10% THD is what high quality system fluorescent systems accomplish.
While better than 0.9 and under <20% meets common requirements, requiring above 0.95 and under 10% reduces the chance of the power quality being met at high flicker, but setting requirements for each parameter is also viable.

Boy, i am sorry you have had so much trouble with leds: fragility, shutdown from overtemp, burnout, flicker, drivers with horrible thd......I havnt had a single one of these problems and all my customers have been super happy.

When LED industry figured out LEDs were failing in reasonable use, they started experimenting with a design that would routinely activate thermal fold back. They end up not meeting specifications simply from not stating many ordinary applications will activate involuntary dimming. Don't get me wrong. Thermal fold back is a good technical feature but it encourages dishonest marketing.

Thermal foldback allows cold storage lighting fixtures rated to operate at full performance under 50F to continue operating at reduced performance instead of burning out or cycling on thermal trip.

Lunera's 20,000 lumen retrofit unit folds back to 15,000 lumen at a fairly normal 120F ambient around a closed top pendant. *if* the customer notices and the ESCO schedules a call back at 9AM, it conveniently comes back concerns unfounded because the customer is unsuspecting that LEDs are purposely dimming down to avoid burning out after running all day and ambient reaches the highest in the late afternoon.

As LEDs accumulate the 30% permanent degradation that is allowed before considered end of life, summer afternoon output at 120F ambient of "20,000 lumen LED retrofit" unit would dip down to 10,500 LM.

"this retrofit can only do 20,000 for a few minutes at a time on your site in this climate"

A 12,000 lumen LED hi-bay fixture that is L90 rated would start off at output comparable to nearly entirely degraded 20,000 lumen unit that is pegged out on thermal foldback. If this wouldn't give appropriate lighting level, the 20,000 lm one mentioned earlier wouldn't maintain it either.

L90 life is hrs until 10% light loss, which is what premium LED fixtures are rated in and is more similar to fluorescent. L70 is 30% loss which is only appropriate in comparing equivalency to brand new MH lamp.

It is odd.

At this point the company I work for has installed tens of thousands (yes that many) LED fixtures of all kinds and we are just not seeing the failures.

Oh for sure there are failuresbut no more so than the failures we were seeing with electronic ballasts for fluorescents, non conformance to claimed output is harder to prove than "it shuts off completely".

I don't know if Electric-Lights wife ran off with an LED salesman but he sure seems to have a vendetta against LEDs to the point of ignoring any success stories .

You gotta make it a personal attack? I believe it was you who posted objective before and after wattage input values, but would not or could not do the same to support output provided appropriate light level for the space or that savings didn't come from trimming light level that was far excessive before that could have been addressed without a change out. HID lamps and ordinary grade LEDs have huge depreciation. When energy sales people trim things too close to tolerance band that they are just barely acceptable OSHA lighting level, it would fall out of tolerance quickly. Prime example of this is brand new LED with 30% decay rating installed in winter that is compared against 70% worn out HIDs with yellowed lens.

 

[Jraef]

Current harmonics are a concern INSIDE of a facility only as it relates to your TDD, Total Demand Distortion, i.e. as a percentage of the total available current demand for the entire facility. The amount of harmonic current created by lighting is generally not the biggest thing to worry about in facilities, unless it is the majority of the load for that facility. The quick and dirty rule of thumb I always use is that if <30% of your total load is non-linear*, you don't need to worry about it. So to the initial point, if by changing to LEDs you lower the total demand current, you in effect lower the TDD-I.

* Non-linear loads are anything with a power supply now, including things like VFDs. So all CFLs and other forms of fluorescent lighting and HID lighting are all non-linear. Linear loads are things without power supplies, and with the elimination of incandescent lighting, all that remains is resistance heating and motors running across-the-line.

 

[Electric-Light]

571-3797742

571-3797742

Current harmonics are a concern INSIDE of a facility only as it relates to your TDD, Total Demand Distortion, i.e. as a percentage of the total available current demand for the entire facility. The amount of harmonic current created by lighting is generally not the biggest thing to worry about in facilities, unless it is the majority of the load for that facility. The quick and dirty rule of thumb I always use is that if <30% of your total load is non-linear*, you don't need to worry about it. So to the initial point, if by changing to LEDs you lower the total demand current, you in effect lower the TDD-I.

* Non-linear loads are anything with a power supply now, including things like VFDs. So all CFLs and other forms of fluorescent lighting and HID lighting are all non-linear. Linear loads are things without power supplies, and with the elimination of incandescent lighting, all that remains is resistance heating and motors running across-the-line.

Lighting power demand is significant in non-residential. And for industrial it depends on the process and if the lighting panel even comes from the same service. ANSI standards mandated <32% THD for ballasts for a long time. When you say "30% of load is non-linear" are you talking about sum of demands from equipment fitting in "non linear" category or the actual harmonic current relative to fundamental current?

Non-exotic, common T5/T8 and CFL commercial fluorescent ballasts achieve a near unity (around 0.98)PF and THD <10%. Also going along with this trend are desk computers that use elective EPA/EnergyStar power supply and server equipment. This topology uses voltage waveform tracking to mirror the current waveform around the voltage waveform. The front-end provides regulated DC power to rest of the ballast/power supply.

The current waveform is in phase with the voltage and it will be sinusoidal but a decent resolution scope will show the sine is drawn in a slightly textured twine instead of a smooth line. But these power supplies are close enough to sine wave to calculate wattage by volt x amp using a multimeter. So it's not perfectly linear, but very close.

NEMA Premium ballasts with a good front end have THD under 10% while many older passive PFC design can get closer to 20%. 200 3 lamp T8 fixtures would therefore be about 15.5kVA, 15.3kW while a total harmonic current for that set of fixtures would be around 5-5.5A (277v circuit).

The basic integral ballast CFL and residential electronic ballasts use passive diodes and capacitors (I think you people call it DC link caps) to get DC and operate the rest of circuits from DC. PF is about 0.6. THD of one such ballast can be over 100% (because of the way math formula works...) and a quantity of them together creating harmonics exceeding 60% rich in triplets. One way of raising PF with corresponding drop in THD is to raise the output DC bus ripple which reduces light quality through increasing flicker. This is what a lot of cheapo LED ballasts do to get >0.90 PF.

>0.95 <15% tends to compel active PFC which brings voltage regulation and flicker reduction alongside.

Line voltage variations affects HIDs a lot. CWA type less so, but tends to be less efficient in general.
Common T5/T8 ballasts and very expensive LED fixtures use active PFC which removes the effect of line voltage variations from power use which can be significant as existing basic HID ballasts use 10% higher watts(and overdriven lamps) at +5% line voltage compared to dead-on.

The basic >0.9, <20% LED retrofits are quite likely to have more flicker than magnetic ballast fluorescent and may not have voltage regulation.

So simply requiring >0.95, <15% or <10% THD with LEDs makes it very likely that you'll get the same flicker and exceptional voltage regulations as it implicitly infers corner cutting is disallowed without having to draw in every corner.

Active PFC intro: http://www.vishay.com/docs/49935/49935.pdf