What causes photocells to fail so often? Why do they fail closed (day burning) and not open (non functional)? I know their are ballasts to prevent cycleing, do they have them for day burning too. In a dusk-to-dawn system, the ballast should shut off and lockout if it gets power for 3 days strait.

From consequence perspective it is prefered to have the lantern failing to dayburning than remaining dark. There are two reasons behind this preference:
First it means it still lights at night, so the prime functionality (so why it is there in the first place) remains. It is only the secondary functionality (= do not use unnecessarily too much electricity), what suffers.
Second reason of that preference is, it is way easier to spot a dayburner during the normal shift hours of the maintenance crew and fixing that type of failure is not that much urgent, so the repair could be better planned in, so it gets cheaper.
Of course, there are plenty of other failures, whose end up in not lighting lantern, but just because there you can not choose (well, without extra expenses), but choice to prefer the photocell would fail closed does not cost anything.


The technical reason behind the failure of the electromechanical photocells is mainly the fact it is a NC type of relay (both magnetic and thermal), so it turns ON when the control circuit is not energized.
And what uses to happen is the heater or coil either directly interrupt (corrosion eating up the thin wire,...) or partly short circuit and the elevated current then blow the remaining section, so again it ends up not energized.
Other failure modes are degraded CdS photoresistor means the resistance is increasing, corroded connections in the control circuit (again the relay not energized) and many others.
So all those failures means the relay contacts stay closed, hence the dayburning.
ANd as normally the dayburning is a prefered failure mode over the dark light, there is no effort in correcting it in the ballast or so (unlike the cycling - that leads to complaints and to more likely arctube rupture).

Thank you for the great explanation. I would prefer failure over day burning. I just hate them, its one of my pet peeves. When I see a day-burner, I see a big neon sign saying "We don't care about the environment and we're to rich and lazy to fix this light." There are several MV day burners on private property around here. The first dayburner owned by the power company (FirstEnergy) around here started about a month ago. I reported it 3 days ago and yesterday when I drove past it (during the day), it was off.

I figured they got it fixed. Today, I just got a call from the POCO saying they could not find it. I don't know how they got my phone number, because I did not leave it on the report. Anyway, I figured if it had been dayburning for a month, it wouldn't randomly stop. So I thought maybe it started cycling and was just out when I drove past. But I drove past again latter and it was still out. Yesterday was about 85F and sunny. I am wondering if the ballast overheated and failed. And since the workorder was for a dayburner, they did not find it.

This is very odd. Would an 85F and sunny day kill a dayburning HPS cobra head?

Quite often I find that moisture gets inside and damages the photoresistor plate and stops conducting in light...which is needed to energize a normally closed relay or heat up a thermal contact to open and shut off lamp. There are other causes as well but corrosion from water seems to be one of the biggest factors.

Voltage surges can also damage photocells. My neighbour had one of 120V and the neutral broke in the power pole, sending 240V everywhere. When the power company repaired and reconnected power, it was a day burner.

With newer electronic photocells, a simple power outage is sometimes enough to make an inductive kickback from the ballast to damage the electronic components resulting in a day burner.

I see a big neon sign saying "We don't care about the environment and we're to rich and lazy to fix this light".

I understand what you mean, but there is one more thing you forget to take into account: Fixing the light mean having there a bucket truck or so, so a 5..7 ton vehicle. If it is dark failure, they have to respond way quicker, that means sending the truck there separately, a trip 10's of km long just for that light. And that means not that little pollution from the truck. If it fails dayburning, you may postpone the repair till the truck will travel close by anyway, so you save most of that pollution from the truck. When we are talking about dayburning up to about a week, the truck is still worse...

The same (although not the topic here) is spot vs group relamping. Many people believe the group relamping is an enormous waste, because "good lamps are wasted", but they overlook the simple facts those lamps have just about 1/4 of their life left (about an average for the the HPS) and the most important, with spot relamping you have one truck trip per lamp, but with the group relamping you have just one trip and many lamps fixed at once.

So the "laziness" is very frequently just to not generate the unnecessary waste elsewhere...

Of course, that does not apply for a dayburner left like that for months...

What happens with some photo cells is that water seeps in and corrodes the relay or other components. My other theory is that the actual photo resistor wears out and goes open circuit (or closed). The lens of the unit becomes cloudy from age, blocking more and more light out as well. The contacts could also corrode very quickly, especially in areas near oceans or other large bodies of water. My photocell on my security light was damaged from water seepage, so I just made it into a shorting cap.

A failed photocell is the most likely cause of the day burning HPS lantern that we can see from our kitchen window. 

But we are scared about reporting it, as they will almost certainly replace it with an LED lantern instead of changing the photocell, which would take mere seconds to do. 

A failed photocell is the most likely cause of the day burning HPS lantern that we can see from our kitchen window. 

But we are scared about reporting it, as they will almost certainly replace it with an LED lantern instead of changing the photocell, which would take mere seconds to do. 

I'd just let it burn until it dies lol

I'd just let it burn until it dies lol

Then it'll get replaced sooner rather than later, it won't last as long dayburning as it would if it were switching normally