[ChainOfFools]

I don't see what's wrong about OP's use of the term in the context they are using it. In the context given the number of days in the denominated unit is 1. Which means as dividend or factor it is going to give you the same result. Again in this context watts per day is much more intuitive for most people too reason about.

[hollerith]

I don't mean to be rude, but anyone who thinks watts/day and watt days are ever interchangeable will have severe problems reasoning about anything electricity-related or energy-related.

It is akin to thinking that "2 apples" and "an apple divided by 2" are interchangeable because both expressions involve the concept of an apple and the number 2.

[mextrezza]

> over 170,000 terawatts of solar energy every day

i'd definitely rewrite it myself, but it's also a correct way to specify that there are no days of the week, year, or whatever (solar cycle) in which the terawattage is below 170k. Not very intermittent, is it!

[ChainOfFools]

I think we may be talking at cross purposes. I specifically used the number one because it behaves like a unit here. unlike 2 or any other number, 1 is also the standard 1D unit vector, so 1apples is indeed the same quantity as apples/1, but because it is a unit we usually imply its presence rather than express it explicitly as above.

watts/unit thus seems fine to me, whatever the unit may be, even if it itself is derived from time. watts per day would just work out to joules/second/1/24*60*60, making 1 watts per day a derived unit that expresses joules/84600 seconds, or an instantaneous rate of one 84600th of a joule.