Neutral harmonics on lighting circuit with 'hi frequency' ballsast?

[tortuga]

Greetings group
For a 4 wire lighting circuit powering a HID lighting load, that uses hi frequency ballasts, do any of you up-size the neural for harmonics?
Circuit size is 20Amp 4 wire 480/277. Load is about 9500 VA and with the continuous factor ~11875 VA (or ~15A)

Here is an example ballast:

Gavita Pro 1000 DE 277-347 | Gavita – Retail

Gavita Pro 1000 DE 277-347

gavita.com

EDIT: I should add the balasts are all 277V L-N loads

 

[synchro]

The spec sheet says the total harmonic distortion is less than 10%. Even if all of that distortion is at triplen frequencies, it's still going to have a relatively small impact on the neutral current.
I would lean toward using a full size neutral but not an up-sized one.

 

[tortuga]

Thanks Syncro! Never thought to look for THD duh!

 

[electrofelon]

Greetings group
For a 4 wire lighting circuit powering a HID lighting load, that uses hi frequency ballasts, do any of you up-size the neural for harmonics?
Circuit size is 20Amp 4 wire 480/277. Load is about 9500 VA and with the continuous factor ~11875 VA (or ~15A)

Here is an example ballast:

Gavita Pro 1000 DE 277-347 | Gavita – Retail

Gavita Pro 1000 DE 277-347

gavita.com

EDIT: I should add the balasts are all 277V L-N loads

I'm late to the party, but thought I would mention:. I have metered not those exact ones, but very similar, dimlux 1000W DE 277 volt, and with 4 on each leg of a MWBC (16A) I got about 2.5 amps on the neutral.

 

[Besoeker3]

There was an issue with this in a very prestigious hotel, The Burj Al Arab. Harmonics exceeded the neutral current. As usual I was asked to look at this at the eleventh hour. The solution was harmonic filters - the 5th, 7th, 11th, and 13th.

 

[Flicker Index]

Greetings group
For a 4 wire lighting circuit powering a HID lighting load, that uses hi frequency ballasts, do any of you up-size the neural for harmonics?
Circuit size is 20Amp 4 wire 480/277. Load is about 9500 VA and with the continuous factor ~11875 VA (or ~15A)

Here is an example ballast:

Gavita Pro 1000 DE 277-347 | Gavita – Retail

Gavita Pro 1000 DE 277-347

gavita.com

EDIT: I should add the balasts are all 277V L-N loads

The harmonics you're concerned with originates from the front-end of the ballast. CBM certified ballasts for commercial use aren't allowed to have more than certain levels of harmonics per ANSI standards, specifically to avoid the triplet issues in the neutral in wye systems. Generally a wide input range ballast (120-277, 277-347, etc) have an active PFC front which keeps harmonics pretty low. A kW class HID ballasts are basically unheard of outside of grow operations though and the common issues these things run into is related to interference with licensed radio spectrum when they're used in a large quantity.

 

[phaset]

The harmonics you're concerned with originates from the front-end of the ballast. CBM certified ballasts for commercial use aren't allowed to have more than certain levels of harmonics per ANSI standards, specifically to avoid the triplet issues in the neutral in wye systems. Generally a wide input range ballast (120-277, 277-347, etc) have an active PFC front which keeps harmonics pretty low. A kW class HID ballasts are basically unheard of outside of grow operations though and the common issues these things run into is related to interference with licensed radio spectrum when they're used in a large quantity.

Echoing Flicker Index on all of this,

kW class SM ballast with silicon (not GaN) switch devices is going to have a core switching frequency between 20-100kHz. When doing conducted emissions analysis at the lab, you'll see something near the core switching frequency, something that corresponds to the clock edge, and then possibly some mixing components at frequencies where parasitics from the board, filter caps, and common mode filters interact. All of this is very, very unlikely to correspond harmonically to the line frequency. And the clocking of multiple units operating together is not synced in any way to the mains freq.

 

[tortuga]

A kW class HID ballasts are basically unheard of outside of grow operations

Not just those operations, LOL
Although they spurred the development of these products and made these popular in all greenhouse agriculture around here.
I was questioning the efficiency of these and was told that these beat LED hands down lumens per watt wise.

 

[Flicker Index]

Not just those operations, LOL
Although they spurred the development of these products and made these popular in all greenhouse agriculture around here.
I was questioning the efficiency of these and was told that these beat LED hands down lumens per watt wise.

I meant to say kW class electronic ballasts are almost never used except for grow-ops where ballasts are within a reach of a simple ladder and they live a relatively comfortable life. Fluorescent electronic ballasts max out at about 200W in common applications, which is the four lamp F54T5HO. They're popular in grow-ups, because they're a lot cheaper to produce up-front and drastically cheaper to ship from China.

There were electronic ballasts for ceramic MH for apparel and produce retail lighting but those too are almost always below 200W and usually under 100W lamp power. GE at one point made UFO shaped electronic HID ballast that fits above the bell jar for 400W HID, but they had such a short market life. I think they had reliability issues. They don't cope well with dust, temperature, condensation and power quality in industrial environment.

BTW, upsizing the neutral was the thing for call centers and such back in the days when computer power supplies had a power factor of about 0.6. Remember those computers with the 115/230 switch on the back?

 

[Besoeker3]

BTW, upsizing the neutral was the thing for call centers and such back in the days when computer power supplies had a power factor of about 0.6. Remember those computers with the 115/230 switch on the back?

That was displacement power factor rather than harmonics.

 

[Flicker Index]

That was displacement power factor rather than harmonics.

um, no. Classic SMPS never had displacement PF. You're talking about motors.

 

[Besoeker3]

um, no. Classic SMPS never had displacement PF. You're talking about motors.

No. I mean UPS and SMPS power supplies. It isn't power factor.It is harmonics. Not 0.6 pf.

 

[Flicker Index]

it is power factor. Power factor can only be defined one way. It is the real power/ (Irms x Vrms). Low power factor caused by non-sinusoidal load has negligible phase shift (displacement) but rich in harmonics.

 

[Besoeker3]

it is power factor. Power factor can only be defined one way. It is the real power/ (Irms x Vrms). Low power factor caused by non-sinusoidal load has negligible phase shift (displacement) but rich in harmonics.

Non-sinusoidal harmonic distortion.

 

[Besoeker3]

"Power factor provides a measure of the efficiency of an electrical system. True power factor consists of two terms: displacement power factor and total harmonic distortion."

 

[Flicker Index]

Displacement (cos phi) is basically irrelevant with SMPS (computers and electronic ballasts).

 

[Besoeker3]

Displacement (cos phi) is basically irrelevant with SMPS (computers and electronic ballasts).

Agree. Power factor is irrelevant for displacement power factor for this application.

 

[Abdel Rahman Mohamed]

Shall we consider the DPF in neutral sizing?

 

[Flicker Index]

Shall we consider the DPF in neutral sizing?

Cos phi is not really that relevant for electronic loads and displacement power factor doesn't do funky thing with neutral current adding up. It is DISTORTION power factor that does funky thing with neutral. Neutral current adding up to high levels was a common theme with 1980s and 1990s IT equipment.

https://www.ecmweb.com/content/article/20889441/how-to-solve-neutral-overload-problems (this article was written in 2001, 20 years ago... when 1990s IT equipment were in abundant use. Active PFC was rare back then. The technology that was common back then was effectively outlawed by EU and with globalization of power supplies so that same units can be shipped globally, everything is built to meet IEC requirements.

Then the last culprits used in quantity that used the rudimentary bridge rectifier input were the CFLs which are becoming phased out too.

While this does not discuss neutral, read the below for fundamental phase shift vs distortion.
See https://cache.industry.siemens.com/dl/files/748/51402748/att_88178/v1/faq_51402748_e.pdf for the dif

 

[Besoeker3]

Cos phi is not really that relevant for electronic loads and displacement power factor doesn't do funky thing with neutral current adding up. It is DISTORTION power factor that does funky thing with neutral. Neutral current adding up to high levels was a common theme with 1980s and 1990s IT equipment.

https://www.ecmweb.com/content/article/20889441/how-to-solve-neutral-overload-problems (this article was written in 2001, 20 years ago... when 1990s IT equipment were in abundant use. Active PFC was rare back then. The technology that was common back then was effectively outlawed by EU and with globalization of power supplies so that same units can be shipped globally, everything is built to meet IEC requirements.

Then the last culprits used in quantity that used the rudimentary bridge rectifier input were the CFLs which are becoming phased out too.

While this does not discuss neutral, read the below for fundamental phase shift vs distortion.
See https://cache.industry.siemens.com/dl/files/748/51402748/att_88178/v1/faq_51402748_e.pdf for the dif

Interesting waveform. Much of what I did were 6-pulse variable speed drives and, of course, rich in harmonics. A few we did were 24-pulse drive systems with very little harmonics. These were usually upwards of 2MVA.