I was thinking about making an HPMV arc tube thinner than the snandard sizes and replacing the argon buffer gas with krypton like it was done with 18, 36 and 58 W fluorescent lamps.
Would it be possible to make for example a 72 W high pressure mercury vapor lamp to replace 80 W ones like 36 W fluorescent tubes replaced 40 W ones?
The only reason I could think of why this wouldn't work is because the starting voltage would be higher so the lamps would need to be pulse start instead of probe start.
Or would krypton have no advantage over argon in high pressure discharges as opposed to low pressure discharges?

In regards to the energy saving mercury vapor lamps that have actually been made, I have not been able to find electrical specifications for the Westinghouse 300W energy saver mercury vapor lamps for North American 400W H33 mercury vapor ballasts as well as the Mitsubishi 375W energy saver mercury vapor lamps for Japanese 400W mercury vapor ballasts. However, I have been able to find electrical specifications for National/Panasonic’s energy saver mercury vapor lamps.

For your reference, here is what I have found out regarding National/Panasonic’s energy saver mercury vapor lamps for the Japanese market:

235W mercury vapor for replacing Japanese 250W mercury vapor:

140V 1.9A

Full power Japanese 250W mercury vapor is 130V 2.1A

280W mercury vapor for replacing Japanese 300W mercury vapor:

140V 2.2A

Full power Japanese 300W mercury vapor is 130V 2.5A

375W mercury vapor for replacing Japanese 400W mercury vapor:

145V 2.8A

Full power Japanese 400W mercury vapor is 130V 3.3A

660W mercury vapor for replacing Japanese 700W mercury vapor:

145V 5.0A

Full power Japanese 700W mercury vapor is 130V 5.9A

940W mercury vapor for replacing Japanese 1000W mercury vapor:

145V 7.0A

Full power Japanese 1000W mercury vapor is 130V 8.3A

I sourced the information from pages 1088-1095 of the 1992-1993 National Facilities and Outdoor Lighting catalog shown below:

https://esctlg.panasonic.biz/iportal/CatalogViewInterfaceStartUpAction.do?method=startUp&mode=PAGE&catalogCategoryId=&catalogId=792140000&pageGroupId=1&volumeID=PEWJ0001&keyword=&categoryID=&sortKey=&sortOrder=&designID=retireOut_sp&designConfirmFlg=


An interesting thing to note about the energy saving National/Panasonic mercury vapor lamps is that they have virtually the same published initial lumen outputs as their full power counterparts while Westinghouse’s 300W energy saver mercury vapor lamps have substantially lower initial lumen outputs compared to their full power counterparts.

I have sourced information about Westinghouse’s 300W energy saver mercury vapor lamps for 400W mercury vapor ballasts on pages 47 and 49 of the 1975 Westinghouse catalog shown below:

https://www.scribd.com/document/112839524/Westinghouse-1975-Large-Lamp-Catalog

More information can also be found in the following post:

https://www.lighting-gallery.net/gallery/displayimage.php?album=lastup&cat=14606&pos=3&pid=265870

How these reduced wattage MV lamps saves energy? Have they different buffer gas or something?

So far, I have not found any information on how they achieve energy saving. If one shows up on Yahoo Japan Auction or Mercari Japan, it would be nice for one of the collectors to use a proxy service to import one into their collection to study its spectrum throughout the warmup process to see if there are any additional buffer gases present.

Making a lamp that draws a different wattage is easy - just change the arc voltage by any means you see fit, be it the arc tube dimensions, fill pressure or fill gas composition

Lower voltage (vs. a standard lamp which arc voltage is not particularly high) will result in lower power draw of the lamp and increased ballast current. How low you can go would be limited by how much overload the ballast can be expected to withstand long term

High voltage within some range will reduce the current provided by the ballast, and reduce the power that way. However, this will very significantly narrow down the range of input voltage variation that the lamp can work with, resulting in a lamp that would go out and cycle for any minor dip in the line voltage, have unstable current that will vary a lot with small changes in line voltage, and probably stability issues over the lamp life. The ballast however will be running cooler with such lamp

The other question is did the lamp actually provide an efficacy upgrade ? Keep the light output of the full power lamp or at least not reduce the output as much as the input power, or just existed as a simple way to reduce the power and light together on existing gear