HID Lamp dimming

[scott]

I have been approached by a company that markets a dimming system (marketed
as energy reduction system) to install their system on a large exterior
lighting for which I am responsible.

The installation is predominantly 150 Watts and 250 watts HPS shoebox
fittings on a carpark.

There are around 500 luminaires involved in one stage, so the potential to
reduce energy consumption is fairly enticing.

The Australian Standard for carpark lighting allows for a reduction of
illuminance levels after a curfew, so the effect of dimming the lamps will
not cause any graet problems with code compliance.

So far, they have not revealed much in the way of the electrical
characteristics of their system, other than to say that they reduce the
circuit voltage at the distribution board, and retain the existing control
gear and wiring system.

I have asked for confirmation that they can attain the same guarantee period
from the lamp manufacturers if the lamps are run on the dimming system, but
so far I have not heard back.

I have had experience with bi-level systems where the multi-tapped inductor
can be switched into a lower power mode, but that of course requires
specialised ballasts in the luminaire itself and 240V up to the primary
side. These systems are widely accepted, and the electrical operation seems
fairly obvious.

Has anyone here had experience with these circuit Voltage controlled dimming
systems? If so, is there any advice you could impart to me?

I have been going in circles with the calculations ( XL=2pifL V=IXL)
etc.trying to analyse the circuit Voltage and current characteristics in
"Voltage reduced" mode. I keep coming up with increased circuit currents,
but not a lot of energy savings. If the circuit current does increase then
taht may cause problems with cable sizing etc.

Any help will be greatly appreciated.

Regards

Scott Forbes

 

[Clive Mitchell]

I have been going in circles with the calculations ( XL=2pifL V=IXL)
etc.trying to analyse the circuit Voltage and current characteristics
in "Voltage reduced" mode. I keep coming up with increased circuit
currents, but not a lot of energy savings. If the circuit current does
increase then taht may cause problems with cable sizing etc.

I recently used a standard low voltage transformer to drop the voltage
to a light fitting by 20V due to the use of a 220V ballast on a 240V
supply. In this instance it was purely to try and prevent the fixture
frying another expensive ballast.

The technique I used was simple. I used a standard low voltage
transformer with the primary wired across the mains and the secondary
wired in series with the mains and the load. Depending on the
orientation of the secondary connections the mains voltage will then
either be boosted by the secondary voltage or dropped by the secondary
voltage. Like an auto transformer with an isolated tap.

In my case I used a 240 to 24V transformer to drop the outgoing voltage
by 10%. The secondary only needs to be rated for the current the
circuit passes so in my case a 12VA 24V transformer was good for 0.5A
which would allow a 240 side load of 120W.

In this application you can either use a single pole changeover
relay/contactor to add the voltage drop in by switching the outgoing
supply between either end of the secondary, or just possibly leave the
secondary in line continually and just power up the primary when the
drop/boost was required. In this case the primary would have to be
treated as live at all times due to the transformer being run in reverse
with the voltage drop across the secondary.

I suppose this arrangement could be scaled up across a few circuits.

Ask if their application is transformer based or uses phase angle
control.

 

[Victor Roberts]

I have been approached by a company that markets a dimming system (marketed
as energy reduction system) to install their system on a large exterior
lighting for which I am responsible.

The installation is predominantly 150 Watts and 250 watts HPS shoebox
fittings on a carpark.

There are around 500 luminaires involved in one stage, so the potential to
reduce energy consumption is fairly enticing.

The Australian Standard for carpark lighting allows for a reduction of
illuminance levels after a curfew, so the effect of dimming the lamps will
not cause any graet problems with code compliance.

So far, they have not revealed much in the way of the electrical
characteristics of their system, other than to say that they reduce the
circuit voltage at the distribution board, and retain the existing control
gear and wiring system.

The versions of these I have seen in the US provide only a
small power reduction, perhaps 5 to 10%. The end user is
told that they can save energy with "no noticeable"
reduction in light, with the clear implication that you can
save energy without reducing light. In fact that are taking
advantage of the fact that your eyes respond logarithmically
to changes in light level while your wallet responds
linearly to changes in power costs.

I would certainly not buy the system until they give you a
data sheet showing the amount of power reduction for the
types of lamps and ballasts you are using.

I have asked for confirmation that they can attain the same guarantee period
from the lamp manufacturers if the lamps are run on the dimming system, but
so far I have not heard back.

In the US at least, lamp manufacturers are reluctant to do
this, not always because they know the systems will reduce
lamp life but also because they have not had the opportunity
to conduct a life test with these devices. Life tests take
a long time, and for HID lamps, a fair amount of energy, if
a statistically significant number of lamps are to be
tested.

I have had experience with bi-level systems where the multi-tapped inductor
can be switched into a lower power mode, but that of course requires
specialised ballasts in the luminaire itself and 240V up to the primary
side. These systems are widely accepted, and the electrical operation seems
fairly obvious.

Don't the bi-level systems reduce the light by 30% to 50% ?

Has anyone here had experience with these circuit Voltage controlled dimming
systems? If so, is there any advice you could impart to me?

I have not, but Terry may have experience with the 5% to 10%
systems used in the US. I assume you are looking for
something closer to 30% to 50% reduction

I have been going in circles with the calculations ( XL=2pifL V=IXL)
etc.trying to analyse the circuit Voltage and current characteristics in
"Voltage reduced" mode. I keep coming up with increased circuit currents,
but not a lot of energy savings. If the circuit current does increase then
taht may cause problems with cable sizing etc.

This would be hard to do since the transformers used in most
HID ballasts are not linear. It would probably be easier to
test one with a Variac or similar device, or the low voltage
transformer suggested by Clive. The only thing these
"energy reduction systems " can do is reduce the ballast
input voltage.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[TKM]

I have been approached by a company that markets a dimming system (marketed
as energy reduction system) to install their system on a large exterior
lighting for which I am responsible.

The installation is predominantly 150 Watts and 250 watts HPS shoebox
fittings on a carpark.

There are around 500 luminaires involved in one stage, so the potential to
reduce energy consumption is fairly enticing.

The Australian Standard for carpark lighting allows for a reduction of
illuminance levels after a curfew, so the effect of dimming the lamps will
not cause any graet problems with code compliance.

So far, they have not revealed much in the way of the electrical
characteristics of their system, other than to say that they reduce the
circuit voltage at the distribution board, and retain the existing control
gear and wiring system.

I have asked for confirmation that they can attain the same guarantee
period from the lamp manufacturers if the lamps are run on the dimming
system, but so far I have not heard back.

I have had experience with bi-level systems where the multi-tapped
inductor can be switched into a lower power mode, but that of course
requires specialised ballasts in the luminaire itself and 240V up to the
primary side. These systems are widely accepted, and the electrical
operation seems fairly obvious.

Has anyone here had experience with these circuit Voltage controlled
dimming systems? If so, is there any advice you could impart to me?

I have been going in circles with the calculations ( XL=2pifL V=IXL)
etc.trying to analyse the circuit Voltage and current characteristics in
"Voltage reduced" mode. I keep coming up with increased circuit currents,
but not a lot of energy savings. If the circuit current does increase then
taht may cause problems with cable sizing etc.

Any help will be greatly appreciated.

Regards

Scott Forbes

You are wise to be suspicious. You should get, at the very least:

-A report on the level of energy reduction that you can expect along with
how much the light will be reduced for your particular lamp/ballast system.
Note that the data for such a report can easily be taken by installing a
test system at your site. A watt meter and relative light readings will
give you the necessary data in an evening.

- An agreement that the company providing the equipment will take over the
lamp warranty. I don't know of any lamp manufacturer that will warranty
their lamps on such a system. It would be good to get a financial report on
the company too. Are they likely to stay in business and do they have the
financial resources to make good a warranty claim?

Forget the calculations unless you have the data for your lamp/ballast
combination. The responses are non-linear as a function of input voltage
and each lamp/ballast is different.

Are there less expensive ways to reduce the system energy? Shutting the
lighting off if there are times during the night when the light is not
needed only requires a time clock; Reducing lighting levels in a
multi-circuit system requires only switches.

Terry McGowan

 

[Clive Mitchell]

Was this an SVC (Sinewave Voltage Conversion) system? These systems
fairly cleanly reduce the output waveform on the dimmer outputs without
using phase chopping (obviously a problem with inductive/capacitive
loads). Generally speaking, they work quite well for a small range
from 100% down to as low as 60% if the system is optimised for their
usage, however, the impact on lamp life is unknown, and I don't know of
any well documented case studies to make up for the lack of science.
SVC dimming may well be the way of the future as the technology costs
come down towards those of phase chopping as it reduces lamp hum,
extends lamp life and manages power factor well. For metal halide? Not sure yet.

Hmm, methinks the simple transformer method might be a lot simpler and
more reliable than some kind of AC to DC to AC again system. I'd
envision a group of transformers with independent power factor
correction if required, and a simple contactor arrangement for switching
in the reduction if it wasn't active all the time.

The lighting load here is quite significant which would suggest that an
electronic system could be liable to be almost unrecoverably expensive
and prone to making very expensive smoke and darkness.

Either way I think I'd limit the reduction to a low level like 10% to
lower the risk of lamp damage through increased electrode work at the
lower lamp temperatures.

 

[SQLit]

I have been approached by a company that markets a dimming system (marketed
as energy reduction system) to install their system on a large exterior
lighting for which I am responsible.

The installation is predominantly 150 Watts and 250 watts HPS shoebox
fittings on a carpark.

There are around 500 luminaires involved in one stage, so the potential to
reduce energy consumption is fairly enticing.

The Australian Standard for carpark lighting allows for a reduction of
illuminance levels after a curfew, so the effect of dimming the lamps will
not cause any graet problems with code compliance.

So far, they have not revealed much in the way of the electrical
characteristics of their system, other than to say that they reduce the
circuit voltage at the distribution board, and retain the existing control
gear and wiring system.

I have asked for confirmation that they can attain the same guarantee period
from the lamp manufacturers if the lamps are run on the dimming system, but
so far I have not heard back.

I have had experience with bi-level systems where the multi-tapped inductor
can be switched into a lower power mode, but that of course requires
specialised ballasts in the luminaire itself and 240V up to the primary
side. These systems are widely accepted, and the electrical operation seems
fairly obvious.

Has anyone here had experience with these circuit Voltage controlled dimming
systems? If so, is there any advice you could impart to me?

I have been going in circles with the calculations ( XL=2pifL V=IXL)
etc.trying to analyse the circuit Voltage and current characteristics in
"Voltage reduced" mode. I keep coming up with increased circuit currents,
but not a lot of energy savings. If the circuit current does increase then
taht may cause problems with cable sizing etc.

Any help will be greatly appreciated.

Regards

Scott Forbes

the others have guide you to the some answers or non answers.
I on the other hand ask this question. Lights burn at night, which means
off peak energy consumption/rate. Cutting the load in half may not dent your
utility bill on a plant that size. Off peak electricity is way less than on
peak where I live.

For security lighting I would consider moving to low pressure sodium less
wattage for the same lumens. Or the switching idea presented by someone. A
motorized circuit breaker that shuts off a circuit of lighting and a time
clock seems like a better plan.

Watchkeeper makes some programmable controls that I have used here in the
states. Nice folks and they support what they sell. Less expensive than the
ole GE LV system.
http://www.pcilightingcontrols.com/soft_wk.htm

Give the guys your utility rate and let THEM provide you with the ROI.
Sounds like snake oil to me.

Let us know how it shakes out.

 

[scott]

Thanks to all who have responded - your time is much appreciated.

I have a face-to-face meeting on Wednesday with the marketing people, so I
will use much of the information you have collectively supplied as
background.

Stay tuned, I will advise you of the outcome.

Many Thanks

Scott Forbes

 

[Clive Mitchell]

I have a face-to-face meeting on Wednesday with the marketing people,
so I will use much of the information you have collectively supplied as
background.

Stay tuned, I will advise you of the outcome.

Please do. That would be interesting.

I'd really like to know if their system is transformer based. They
might want to hide the operating principle from you with technical
nonsense.

Then again. If you're dealing with the salesmen they won't know how it
works anyway.

 

[scott]

Clive, they have already as good as admitted they have no idea about its
electrical characteristics, but hasten to add that they have plenty of
installations across the world as testimony to the fact that it works! I'm
hoping that the salesman will get on his phone and contact an engineer when
he gets into deep water.

You are correct in your assumption that they are trying to baffle me with
bullshit. It will be interesting to watch just how deep he digs his hole.

 

[Victor Roberts]

Clive, they have already as good as admitted they have no idea about its
electrical characteristics, but hasten to add that they have plenty of
installations across the world as testimony to the fact that it works! I'm
hoping that the salesman will get on his phone and contact an engineer when
he gets into deep water.

You are correct in your assumption that they are trying to baffle me with
bullshit. It will be interesting to watch just how deep he digs his hole.

Let me add one comment that I had intended to put in my
initial reply.

Years ago when I worked for GE I visited the main office of
a company in the US that sells energy reduction systems.
They tried to tell us that GE Lighting had approved their
system for use with GE lamps. I knew that was not correct
so we asked for the proof. They showed us a letter on GE
stationary that said their system was just great! It turns
out that they had convinced a plant manager, or perhaps the
plan's maintenance manager, to install one of their systems
in one of GE's many lighting plants, and this manager
provided this letter of "approval" based only on his opinion
and without any data or approval from GE Lighting
management. His letter constituted a personal statement of
approval for the system but certainly not official GE
validation of the system as implied by the energy reduction
company.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[Victor Roberts]

I posed a question a few days ago that I believe is indirectly related
to your problem. Unfortunately, I got no response to my posting.

I don't know about others, but I can tell you why I didn't
respond.

Your original post asked for confirmation that certain
capacitor values could be used to raise the output power of
a ballast that was designed for a higher line voltage than
you intended to use to power the ballast.

Unless I was very familiar with that ballast, or had the
ballast manufacture's design data available to confirm your
proposed mode of operation, I could not answer your question
with any confidence. And, when working with metal halide
lamps, operation at greater than rated power, which could
result if your calculations were incorrect, could lead to
lamp explosion and property damage, injury to humans or even
death. Metal halide lamps should be run only on ballasts
designed for that type of metal halide lamp.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[Don Klipstein]

I posed a question a few days ago that I believe is indirectly related
to your problem. Unfortunately, I got no response to my posting.

As far as I know, the current method for dimming or changing lighting
levels for an HID is by switching in/out an auxiliary capacitor. This
does not require a special ballast, but does call for a relay in the
housing to toggle the extra capacitor in or out. Some of the ballasts
and associated capacitor information is included in the current Advance
HID Ballast catalogue. I found it in the
2002-2003_Advance_Atlas_Section_4_High_Intensity_Discharge_Ballasts.pdf
on page 4-48 under the section labeled "Recommended Capacitors for
Bi-level Ballast Operation". They list the capacitance values for
lowering the light output, and the associated drop in wattage consumed
by the ballast.

There are many companies that have systems based on this feature, some
which use programmable addressable relays to control individual
ballasts, or banks of fixtures. One such system I found as an example
is described in: http://www.wide-lite.com/Literature/cntrls.pdf. From
your description, I believe this technique may be what you are looking
for.

What I wonder is what is the effect of slightly increasing, or
decreasing the capacitance in a CWA ballast - a 400w Metal Halide in
particular. I was offered an inexpensive high-bay fixture with a
277v-only ballast that I considered using on a 240v feed. I assume that
if this is a proper way to alter an HID output, then that may be a
method to correct for the lower input voltage (240v) I would have to
use.

I see a possible problem: Capacitors that do what you want on one
ballast may not do exactly the same thing on another of the same basic
type and ratings. What happens could vary with brand, model and
production run of the ballast.
Keep in mind that capacitors have tolerances and temperature
coefficients. Also, the ballast's regulation characteristics (degree of
overpowering/underpowering the lamp due to varying line voltage or lamp
voltage drop) can change with use of a capacitor other than what the
ballast manufacturer intended for that particular ballast.
Keep in mind that if current through any ballast winding is above
intended warmed-up value (as opposed to higher values allowable only
during warmup), the ballast could overheat, and not necessarily quickly.
Also watch voltages across capacitors, especially in early warmup - they
could exceed the ballast open circuit output voltage, and vary with line
voltage and capacitor tolerance and capacitor temperature.

I have even seen in some experiments with HPS lamps and experimental
capacitive-inductive ballasts the lamp having conductivity increase as a
half-cycle of AC voltage progresses, making the lamp have some simulation
of an inductor. This effect could vary from one lamp to another and vary
with the lamp's age and condition.
One thing to be wary of with an experimental or modified ballast with
capacitance in series with the lamp is the lamp being overpowered or
merely being run when the lamp voltage is highly excessive due to age/wear
- while with a proper ballast used as directed the lamp may instead cycle
or refuse to run. I have gotten HPS lamps to even have lamp voltage above
the ballast's open circuit output voltage by using series capacitance in
experimental setups - and the arc tube temperature and internal pressure
in that case I expect to be outside the ranges that the arc tube was
intended to withstand! Such experiments I did in my more reckless younger
days, and I had explode an oil filled capacitor about the size of a beer
can in the process!

If you do anything not specifically recommended by the ballast
manufacturer for your particular ballast, then your system is
experimental. That is somethinmg I would run in my basement, and only
while I am around to use the fire extinguisher if necessary - and keep the
lamp in a suitable enclosure in case the lamp explodes, even if the lamp
is a "protected" type. Or wear safety goggles and be willing to endure
burns from hot lamp fragments, and don't have much of anything burnable
around.

If you are going to use an HID lamp for lighting rather than for some
experiment, you may as well get a proper ballast with a lamp it is rated
for and use them as recommended by their manufacturers. If you burn down
a building doing other than this, you could have extra liability - even if
no voltage, current, or other ratings of anything were exceeded.
If the tooth fairy's evil aunt not liking you is the only reason your
ballast or lamp starts a fire, the fire insurance company will give a lot
less grief if the ballast was being "used as directed".

- Don Klipstein ([email protected])

 

[Peter Pan]

Make sure that whatever it is, it isn't a phase controlled system. This
will cause trouble special with HPS. However you can change the power
on a magnetic ballast with one of the following methods:
1) Frequency control (as a motor control)
2) Changing a capacitor in a circuit like a CWA ballast
3) Changing a tap, if the ballast allows for such use, or adding
another inductor to the circuit

I found a company that makes a ballast that reduces the power by 50% of
MH lamps However, they had mentioned that doing it on a HPS ballast
shouldn't be that difficult, though they never had anyone ask for that.

As for a lamps life, it is important that the lamp is started and
stabilized at full power. Otherwise the life will be shortened.