Welcome to this blog, which looks at the potential perils of switching over from HID lighting technologies to modern solid state LED lighting.
It is a commonly made [but incorrect] assumption, that solid state/LED based lighting products should be as robust as old magnetic/HID lighting products. Many of us have heard stories of old magnetic HID lamps catching fire on sites– and still operating [even afterwards], this is a good example of how robust HID gear can be with regards to Ta limits! – Obviously modern LED luminaire’s contain plastic [in their LEDs casing] so would not survive such treatment- But, there are also other equally important differences to be considered.
We advise users to make sure they know what the limitations of solid state/LED based luminaires are -when replacing nearly indestructible magnetic gear on sites which may have environmental and power quality issues which have not been tested in use with solid state lighting before.
Old magnetic control gear technologies [such as HID/fluorescent] are well known for their ability to handle harsh industrial environments- such as sites where the following issues may occur:
Surge and interruptions.
Fast transient voltage spikes.
Excessive current harmonics from heavy machinery [such as motors/arc furnaces /resonating HIDs].
Onsite voltage distortion [caused by heavy machinery]
Excessive current through a neutral line [caused by harmonics] with poor earth bonding- leading to a raised voltage on the neutral [with high voltage harmonic levels]
Imbalanced three phase supplies
High onsite temperatures [possibly combined with power quality issues that also increases internal generated heat in the luminaire]
Lack of access to light fittings in dirty/dusty environments- where luminaire cleaning/servicing is not performed. [Possibly combined with the previous point above]
Sites that use/emit organic solvents/chlorine/dust/liquids/materials that can thermally insulate fittings –especially if they are not regularly cleaned.
Very important note:
It is of vital importance to know what the limitations/differences of the new LED lighting products are with regards to replacing HID based lighting on site with any of the issues above.
To blindly assume that new solid state LED gear will identically match the robustness of magnetic gear in the same environment is to court disaster.
We know for example, that introducing triac based dimming on a noisy AC supply on a heavy industrial site [with ringing affect/current harmonics from HID gear+ harmonics from motors/heavy machinery] would be bad choice, as solid state triac dimmers are more very prone to onsite EMC issues as opposed to old magnetic control gear -the result would be an unstable dimming control with flicker issues. -likewise, when looking to using technologies other than magnetic based control gear on sites with heavy machinery -such as SELV driven LED or AC direct driver LED, we would also consider onsite power quality issues that could affect the particular designs/types of LED products available. It is wise- not to assume, that all LED products us the same design –with the same specification /environment handling [even products using the same driver/LED can differ in thermal performance and EMC]
It should be noted that modern solid state lighting products are rated as C class [ under BSEN 61000-3-2 Limits for harmonic current emissions using less than 16A per phase] -this specifies the allowed amount of current harmonics/voltage distortion the product will emit, and also the allowed degree of voltage distortion the product will happily operate from.
Heavy site machinery, including arc furnaces/large motors create a much higher level of current/voltage harmonics than found in class 1or class 2 pollution environments and can damage class C equipment, if the low voltage power line to the class C equipment is not is not filtered/corrected with regards to harmonic levels being non-compliant to the limitations set by BSEN50160.
One common old site issue, is high excessive neutral currents [ + if earth bonding is poor and the neutral current harmonic level is very high with a noisy voltage present on the neutral ] in such cases ,the rectifier stages of solid state control gear/LED drivers can be switched on/off rapidly by heavy voltage distortion [HF noise]. This creates additional diode inrush currents and also switching noise into the driver- the high level of injected noise into the driver’s rectified DC supply can on some designs -cause interference with the driver’s oscillator stage [affecting output current stability/flicker]. On other designs it could lead to overheating of the HF filter stage on the AC input of the driver.
The result is the driver’s rectifier diodes/AC filter are working harder [more often- increasing heat/reducing driver lifespan] The additional internally generated heat will lower the Ta rating of the luminaire and possibly reduce the LEDs lifespan and L70 performed capability, Note: such heavy industrial sites may have high onsite ambient temperatures and dust levels [ also the luminaires heatsinks may be rarely cleaned ] so a lowering of the luminaires Ta could prove to be catastrophic, the site may think the luminaires failed purely because of temperature issues and not identify the voltage distortion as the main contributing cause.
Many solid state driven fluorescent luminaires may have great difficulty striking up when wired in parallel to a resonating HID lamp or high voltage harmonics from heavy machinery - due to interference of the drivers start-up sequence caused by voltage harmonics/interruptions or low voltage levels, in such cases repetitive failed attempts to strike fluorescent tubes during periods of high onsite harmonics can eventually overheat/wear down/blow the luminaires output semiconductors.
Some solid state drivers employ power factor correction circuity- which senses and adjusts the drivers current consumption dynamically- depending on the phase position of the incoming AC wave- however, if there are high voltage distortion levels present, the correction circuit cannot sense the AC waves phase position correctly– so the drivers power factor sensing is thrown out ,possibly leading to even more voltage onsite distortion –complicating the situation further [depending on make design of driver]
On LED drivers with very low power factors [0.5] – the timing /phase angle of the current flow will be corrupted by onsite voltage distortion- depending on design, this may affect light output stability/power consumption and could cause PWM flickering, as well as a possible breach of current harmonic emissions from the luminaire.
In the past, older robust HID control gear was often used on sites with high harmonic levels [without too many issues other than flickering and occasionally bursting into flames] unlike solid state gear, the old magnetic transformers do not immediately fail through overheating/overvoltage- as mentioned before they can even work when on fire, as they do not have the same thermal protection safety features!
It is worth noting that resonating HID control gear can cause voltage distortion/interruptions that can cause flicker on the HID lamps [ and other parallel lighting] Most modern LED fittings are designed to be mainly flicker free , so when wired in parallel with such resonating HID units ,the LED driver may experience rapid on/off switching/inrushes -while the LED lamps internal electrolytic capacitors will be working harder to reduce flicker -decreasing capacitor lifespan and introducing additional internal heat, thus lowering the LED products Ta rating.
All Class C lighting equipment is intended/designed for use on onsite supplies that are compliant to the limitations described in BSEN50160 this standard [as used by all good power quality loggers] provides a set of parameters/limitations which are commonly expected on industrial sites – when voltage or current harmonic levels exceed these stated limitations – we would also expect to see performance problems with other onsite solid state equipment such as computers/audio equipment.
Other important differences between HID and LED luminaires-
Vitally important note: Dust handling /Cleaning schedule
In comparison, a typical LED luminaire employs a heatsink, and just like any other heatsink – if it is covered in dust/dirt it will cease to cool down the LEDs/driver and they will overheat/fail – on all solid state based products it is vital to keep the heatsink working/clean, otherwise the luminaire could thermally cut-out, or derate [derating is where the light output is reduced to avoid driver overheating] or fail to meet L70 performance requirements.
One long term result of non-serviced heatsinks –could be luminaire output derating, which could cause onsite safety issues with regards to acceptable lighting levels and can lead to a legal dispute.
While a magnetic ballast can burst into flames [if overheated by onsite conditions] a CE certified LED luminaire will derate and eventually thermally shut down or shut off to avoid becoming a fire hazard.
To assume LED luminaires heatsinks will not require regular cleaning [because the old HIDs didn’t] would be a great error – especially on sites with potential high temperatures and high harmonic levels.
Remember- a heatsink cannot do its job if it is covered in thermal insulating dust- the result is always eventual overheating/failure –and replacing the product for another make/type [without introducing a cleaning regime] will simply reproduce the same result – as all high power LED products using a heatsink will fail if they overheat, it is surprising how many sites ignore this basic common sense fact – would anyone buy a car and never service the radiator while expecting it to always work on the hottest days –for years ,no matter how good the car is ,it will overheat/fail eventually.
Good servicing of luminaries heatsinks and diffusers can extend the products life span and L70 performance. Some luminaires are over designed with additional Ta and L70 headroom [to be safe] – if they are kept clean, they may perform well beyond 50,000hrs – so cleaning the heatsinks can be an economic benefit.
We acknowledge that types and levels of dust/dirt/solvents will vary site by site – so it is hard to define a standard cleaning schedule period that would suit all factories, on sites where dirt levels are excessive [and also where temperature is a potential issue] we recommend contacting the luminaire supplier for advice.
If the manufacturing materials used onsite are changed – [changing the dust level/type] it could affect the required cleaning schedule. Likewise, users should be aware of any factory changes which would alter the temperature of the heatsink [such as the introduction of new machines onsite that may emit additional chemicals +heat]
Old glass lamps and magnetic control gear are far more resistant to onsite chemicals such as organic solvents and gases LEDs can be corroded/discolored by organic chemicals – Always check with the luminaire supplier, as to what kind of chemical protection is required for the sites fittings [for example: if a site emits chlorine gas the luminaires LEDs would have to be protected using gaskets/conformal coating – otherwise the LEDs will fail prematurely]
Depending on particular sites – heavy onsite vibration may damage/disrupt some LED luminaires, while not affecting old screwed in HID lamps. Many LED fittings are vibration tested for such use – but some are not, always check if there is doubt.
Very important note- when changing over
Problems can occur, if a site has HID lighting on all three phases, and the fittings are swapped over to LED over an extended period [not all done in one go] and the fittings are only swapped one phase at a time [ so one phase is running LED products while the others are still HID]
By reducing the power consumption/load on one or more phases, the supply will become imbalanced- this can lead to an overvoltage of the LED products and can also introduce/increased current harmonics and fast voltage transient spikes/interruptions- bear in mind, the remaining resonating HID circuits will behave differently as the three phase supplies EMC characteristics are altered – each time a fitting is changed, a new scenario can occur.
The result is a potential multitude of problems which may only occur, when the two types of technologies are combined onsite.
Even amongst LED products, designs can vary greatly to say the least – not all LED drivers are SELV/constant current types [some are AC direct or high voltage LED] some have power factor sensing –some do not. Some drivers may be more susceptible to mains noise/harmonics [generated by HIDs]
Not all LED drivers have the same level of immunity to fast step voltage changes [this is where a sudden 20V change in line voltage causes momentary flicker]
The result can be a site with differing symptoms each time the work is progressed by a few fittings – if different LED products were tried on the same site with resonating HIDs – it could produce a different symptom on each different make/model of LED product, but the cause would be the old flickering HID units emitting non-complaint levels of current harmonics, which may look like they are working fine.
Ideally, all three phases and all fittings should be changed over in one go- to avoid supply imbalance and the unknown interacting harmonic issues which can occur at various stages of change over. However – this is not always practical.
While old sites must ensure the low voltage supplies power quality is complaint to the expectations of EN50160, the power quality will vary- if the swap over involves using both old equipment and new together [it is impractical to monitor constantly –cost wise].So a scenario can occur where the onsite supply is non-compliant for a period [as it is in a half way change-over state]
However, it is hard to say what will happen with each individual site when combining lamps types on a slow changeover– it may be no problems occur, or other onsite machinery being switched on/off [unbalancing the three phase] as well as an imbalance on the lighting circuit causes an even greater issue – But, we can say for certain, if the HIDs are changed all in one go, that the HIDs will not affect/damage the LED products [though, this is not a guarantee of onsite power quality]
A site could switch off the HID lamps, and then run a power quality test [looking for harmonics/dips/ interruptions/ spikes/surges] this would indicate the suitability of the onsite supply when the HIDs are removed .Should EMC problems occur when performing a slow changeover [combining HID/LED] the installers would then know the most likely cause of disruption was the old HID units, and not the LED products or other site machinery.
If half way through a changeover, a contractor experiences performance issues with LED fittings [such as flickering/failure/derating] they may think the LED fittings are at fault and not identify the real cause- they could purchase a different make of fitting and still be facing the same problem .A power quality test would then indicate the cause [but it would have been cheaper/quicker to just change the remaining HID lamps]
We would be concerned that damage could occur on installed LED fittings if left operating on supply non-compliant to EN50160 [due to the use of old HID gear in parallel ].in all cases we recommend an onsite power quality analysis if in doubt ,and to avoid mixing old and new lighting technologies together on the same circuit.
We hope this blog is helpful to installers/specifiers in avoiding such potential onsite pitfalls.