The high lumens per watt range of the LEDs are only possible in a magical environment with the lack of many factors, once you you add the ballast losses, the weather, the humidity level, how fast the heat is transferred among others you will end up with nothing more than 100 lumens per watt. The big problem is the ballast, because as more powerful more losses in transistors and resistors, and in most retrofits and integrated fixtures the ballast is at the very side of it probing heat to it, also the humidity of the air will affect how fast the heat dissipates to the air. Also if you put many in series and parallel you will put more stress into the board and it will heat up. Also, the whole thing will suffer of the thermistor effect as it heats up, decreasing the efficiency of the chips and the ballast.

If you light up a single chip from a lab bench power supply at 25 degrees 50% humidity you will get great numbers, once you bring it outside and with the ballast the numbers will go down.

The high lumens per watt range of the LEDs are only possible in a magical environment with the lack of many factors, once you you add the ballast losses, the weather, the humidity level, how fast the heat is transferred among others you will end up with nothing more than 100 lumens per watt. The big problem is the ballast, because as more powerful more losses in transistors and resistors, and in most retrofits and integrated fixtures the ballast is at the very side of it probing heat to it, also the humidity of the air will affect how fast the heat dissipates to the air. Also if you put many in series and parallel you will put more stress into the board and it will heat up. Also, the whole thing will suffer of the thermistor effect as it heats up, decreasing the efficiency of the chips and the ballast.

If you light up a single chip from a lab bench power supply at 25 degrees 50% humidity you will get great numbers, once you bring it outside and with the ballast the numbers will go down.

Thea "magical environment" is a decently designed device. That means chip temperatures about 60..80degC, that is where mose decent LED makers use as reference temperatures for rating. Dont expect that from the "suupeer qualitee only from aleebabaa", there you would be indeed lucky to find something barely reaching 100lm/W for CRI60.

With minimalistic chip at minimalistic currents and cryo temperatures the LEDs are above the 200lm/W ballpark, but that is, indeed of no direct value for real life (more an indication how far the development may go eventually, if no fundamental limitation would be hit).

The ballast losses are in the 20% ballpark for the super compact filament designs (where is no room for anything else than a linear CCR device) with limited power, but the streetlight ballasts (so what is really used with decently build streetlights) are less than 10%.

And once you add typical LPS ballast losses, you get around 150lm/W for just a monochromatic light near centre of the human sensitivity and with no color at all. And only for the highest wattage lamp, usable maybe for airports (of the past, as now certain minimum color rendering is required for safety reasons) or maybe a rail yard (however there the zero color rendering means all signage must be illuminated by itself, so about the same extra power consumption there)
Yes, there is no direct competitor product (monochromatic light in the yellow-green eye sensitivity peak area, so where relatively low efficiency suffices to reach high efficacy) in LEDs, but there is not much use for such light source either. Maybe except hevy fogging places, most other had shown need for at least some color perception (distinguish different materials that may have fallen off from something breaking up, no need for traffic sign illumination while stil being well readable even outside of the headlight beam,..,).

The thing is that most municipalities or manufacturers wil go for those cheapo "suupeer qualitee only from aleebabaa" rather than actual quality LEDs in the 150 lumens per watt range, so in most cases best you can get is 110lm/w, and because the light efficiency decreases with the time the average will be less than 100lm/w. Also, in cheapo and probably some decent fixtures too they squeeze the LEDs to the moon, so the LEDs are working in the edge or overloaded, so if the tests are made at 60-80 degrees at the end of the day they will operate at 90-100 degrees. Leaving a redundant gap from the maximum work power will increase the life and the efficiency too, but that's economically non-viable, because it would need to increase the amount of chips per fixture or decrease the power per fixture but with the same amount of chips, which is likely to never happen.

Also the LED changeouts are just pointless in my opinion, because they won't perform significantly better than HID or LPS/SOX, less CRI is safer, but if you decrease the CRI of an LED the blue light increases, which is bad news for people trying to sleep or for the eyes of drivers, and they won't last as long as HID, also LEDs tend to blink at EOL, that can trigger seizures in a sensitive person, which can lead to a salty demand for who made the changeout, also high CRI streetlights are the paradise for drvg consumers, which can increase the violence and crime.

In my opinion the LED changeouts are just for money. You can save a lot of money in maintenance because you leave the fixture there and when it fails you just replace it with a new one and it ends up being cheaper than a regular maintenance to a HID of a relamp. Also the government might pay off part of the cost for supporting the "green energies", as I said before.

The thing is that most municipalities or manufacturers wil go for those cheapo "suupeer qualitee only from aleebabaa" rather than actual quality LEDs in the 150 lumens per watt range, so in most cases best you can get is 110lm/w, and because the light efficiency decreases with the time the average will be less than 100lm/w. Also, in cheapo and probably some decent fixtures too they squeeze the LEDs to the moon, so the LEDs are working in the edge or overloaded, so if the tests are made at 60-80 degrees at the end of the day they will operate at 90-100 degrees. Leaving a redundant gap from the maximum work power will increase the life and the efficiency too, but that's economically non-viable, because it would need to increase the amount of chips per fixture or decrease the power per fixture but with the same amount of chips, which is likely to never happen.

Also the LED changeouts are just pointless in my opinion, because they won't perform significantly better than HID or LPS/SOX, less CRI is safer, but if you decrease the CRI of an LED the blue light increases, which is bad news for people trying to sleep or for the eyes of drivers, and they won't last as long as HID, also LEDs tend to blink at EOL, that can trigger seizures in a sensitive person, which can lead to a salty demand for who made the changeout, also high CRI streetlights are the paradise for drvg consumers, which can increase the violence and crime.

In my opinion the LED changeouts are just for money. You can save a lot of money in maintenance because you leave the fixture there and when it fails you just replace it with a new one and it ends up being cheaper than a regular maintenance to a HID of a relamp. Also the government might pay off part of the cost for supporting the "green energies", as I said before.

Keep in mind that SOX and LPS are not the exact same thing and that SOX is a type of LPS lamp.

Keep in mind that SOX and LPS are not the exact same thing and that SOX is a type of LPS lamp.

I know, I put it like that because their efficiency is not too different, but I will correct myself and I will keep it as just "LPS".

I know, I put it like that because their efficiency is not too different, but I will correct myself and I will keep it as just "LPS".

LPS is just a broad lamp category while SOX is specific.

The thing is that most municipalities or manufacturers wil go for those cheapo "suupeer qualitee only from aleebabaa" rather than actual quality LEDs in the 150 lumens per watt range, so in most cases best you can get is 110lm/w, and because the light efficiency decreases with the time the average will be less than 100lm/w. Also, in cheapo and probably some decent fixtures too they squeeze the LEDs to the moon, so the LEDs are working in the edge or overloaded, so if the tests are made at 60-80 degrees at the end of the day they will operate at 90-100 degrees. Leaving a redundant gap from the maximum work power will increase the life and the efficiency too, but that's economically non-viable, because it would need to increase the amount of chips per fixture or decrease the power per fixture but with the same amount of chips, which is likely to never happen.

Also the LED changeouts are just pointless in my opinion, because they won't perform significantly better than HID or LPS/SOX, less CRI is safer, but if you decrease the CRI of an LED the blue light increases, which is bad news for people trying to sleep or for the eyes of drivers, and they won't last as long as HID, also LEDs tend to blink at EOL, that can trigger seizures in a sensitive person, which can lead to a salty demand for who made the changeout, also high CRI streetlights are the paradise for drvg consumers, which can increase the violence and crime.

In my opinion the LED changeouts are just for money. You can save a lot of money in maintenance because you leave the fixture there and when it fails you just replace it with a new one and it ends up being cheaper than a regular maintenance to a HID of a relamp. Also the government might pay off part of the cost for supporting the "green energies", as I said before.

The problem is comparing $200 fixture requiring extra $40 lamp with $50 complete fixture, the later very likely wont outperformthe first.
But the question is, if it really needs to.
At many places higher power was installed just because it seemed to cost nothing extra, when the 50W cost the same $140 as the 250W (so really what was a 125W MV in the 70's very often became 250W HPS over the years) and the electricity becomes available for free (that was the case for many municipalities with their public lighting tarriffs). But now these tarriffs are getting revised and it starts to be pretty expensive for the towns to operate 250W where 70 or 50W of the same HPS is enough.
And with that, they may choose more expensive 35W good quality fixture for $300, or a bit less efficient 50W one sold for $50. The thing is, the 15W difference wont make for the $250 over the lifetime. The fact the cheap one wont last that long so will need more frequent replacements, which then brings the total cost about the same. In either way, it is way cheaper than running the 250W any longer...
Well, makers of both knew that, so that is why their product are priced that way in the first place (so it cost you the same either way).

The problem here are the various "anticorruption laws", which mandate public officials to use the cheapest offering fulfilling the formal requirement set for the bidding. And if those officials use even "past experience with the supplier" as an advantage/disadvantage for some supplier, they are facing nearly jail time for "interfering with free competition" and "neglect of public funds" and so on. Even putting requirement for extended warranty into the formal requirement was used to pu those officials down, for "inadequate boosting of the price". Really these "protection" laws have to be adressed first, as usually these are the reason behind the cr@p that does not work being purchased for public money.

There has been quite many cases of public funds frauds due to the LED changeouts. The one who have the task of choosing the supplier and importing the streetlights choose a fake company as the supplier, then the employee will choose a premium or expensive product, then the fake company will emmitt the total bill and the superiors will approve the fund to spend. Then the fake company will import crappy junk streetlights for dirt cheap or there will not be streetlights at all and the involved people will keep the rest of the money till they get caught, if they do.

Ok people, we seem to be straying from the topic under discussion here, this is about how efficient HID lamps are, not the reasons (political or otherwise) why they're being replaced.

Back on track please, we don't want to be locking another topic.