I'm thinking of something like HPS with a PCA arctube. They might be easier to start, as they have lower melting points.

I'm thinking of something like HPS with a PCA arctube. They might be easier to start, as they have lower melting points.

But then you end up with the need of a pulse start and higher OCV margin, essentiaally a HPS, just with different elements replacing the sodium. Dunno what the overall color and efficacy would be, but I bet such ideas are nothing new, because these elements are well known to the HID making community, so they are not used in this form because they are just not viable materials for such lamps...

How does a shorter, high pressure arc change the electrical characteristics? By how much does lumen output increase?

It is all along wht is depicted in Fig23.
The electrical characteristics (arc voltage at a given arc current) have to remain the same, because he lamp is supposed to electrically match the standard.
So when you want to increase the arc loading without increasing the power, you have to make the arc shorter.
But when keeping all other variables the same, a shorter arc means lower voltage drop, hat you do not want.
So to increase the voltage drop back to where it is supposed to be with the given wattage, you have to boost the mercury pressure.
And to boost the pressure, you need higher operating temperature.
And the higher temperatures and pressures is, what accelerates the aging.

Because with modern factory you may reach better manufacturing consistency, you have less early failures.
With less early failures, you may get a room to sacrifice the median life when still maintaining the 2% failure rate at 16khours (to get 4 years of service interval, that is, what the public lighting market requires).
This extra room allows you to run the materials harder, so to get some extra efficacy from your product so gain some advantage over competitors.
And that is, what happened with MV's in their last two decades...

The electrical characteristics (arc voltage at a given arc current) have to remain the same, because he lamp is supposed to electrically match the standard.

But lets say you didn't. Lets say you had the power to design the auxiliary gear or halogen tube anyway you wanted.


In my world I am willing to sacrifice life for efficiency.

But lets say you didn't. Lets say you had the power to design the auxiliary gear or halogen tube anyway you wanted.

In my world I am willing to sacrifice life for efficiency.

In any way you push the pressure/temperature as high as you are comfortable with regarding the reliability you need.
Then you have to take into account which type of gear will be used, in other words what is the mains voltage for the application.

Simpler case:
For 230V your ballast of choice will be just a series choke, because of the least losses compare to other ballasts.
That implies your arc voltage would have to be around 120V (higher would mean too little OCV margin, so unstable arc, lower means higher current, so higher electrode and ballast losses).
This then dictates the electrical characteristics of the final lamp (and why do you think it is the same as standard lamps).
Then you design the arc length to match this arc voltage in the desired fill pressure.

A bit more complex case:
You have 120V mains, what clearly does not allow any reasonable efficacy on just a simple series choke (would need 60V or below arc). The designed arc voltage will have to be then the compromise between electrode dissipation (the low arc voltage would mean high percentage would be the nonluminous electrode drop, so low overall efficacy) vs the efficacy suffering from too low arc loading (higher arc voltage means longer longer arc, with the same power that means lower arc loading, so lower efficacy from the luminous part). So you go for the arc voltage/arc length corresponding to the maximum efficacy.
Then you have to design the ballast accordingly (the current, OCV,...).