There's no answer for that.
Not every plant differentiates by experience. (This really p.o.'s the more experienced ones.)
There is a vast regional disparity between pay rates.
Those plants which do have progressive pay rates usually cut off at seven years. I don't know why this number is significant. Perhaps someone really believes that you can learn all there is to know in 7 years and coast for the next 20 on that. OR. maybe anything that happened over 7 years ago is irrelevant to current methods. Both assumptions are flawed.
There are also (as you should know) three widely different definitions of the word "average". There is the mean, the mode, and the median. Anyone can use whichever one of those suits his own argument. Not having worked at all of the plants in the world, I don't personally have enough data points to calculate any average.
Most importantly, I do not CARE what the average is. I don't decide on what is an acceptable pay rate based on any statistical formula. I make a subjective judgement based on the offers in front of me at any given time. I usually pick whatever job is going to be better for my family's interests overall. (That means that I make short-term decisions with my eye on the log-term implications.) Translate that to read: "where can I make the most money in the shortest time without killing all my other options?" If an outage is scheduled so that I could have worked two others, but had to miss them both to do this one, I better bring home as much as I would have for BOTH of the other two or I'm going to choose the other two. If I can get 3 weeks of 84's or 3 weeks of 60's, I'm going to do the 84's. Sometimes higher pay makes up for fewer hours or shorter duration, but you put any combination of low pay, short hours, and/or short duration together and it will probably go to the bottom of my list.
As a side note, you don't see this in some specialties. Some contractors (not HP) send people to whatever outage needs the people. They pay them essentially the same no matter where they assign them, and give them all the hours they can. Then, they send them right to where they need them next. Employees don't hold the sites hostage over pay rates, hours or working conditions, and they don't have to. These jobs are always staffed. When's the last time you heard refuellers have discussions like the one here?
Palo Verde shouldn't care about the average either, at least not beyond using it as a reference point. What they should care about is how much it will cost to get what they want - namely, the required number of qualified techs. Arbitrarily offering incremental pay rates based on returnee status, etc. isn't going to guarantee this. Essentially, they should be able to staff every outage with the exact same group of techs, save for those who cannot return because of prior committments.
The fact that they need to pay people more to come back says something about the way they treated them the first time. Why would a tech not want to return to a good job without being bribed?
If you're getting a really good percentage of returnees, then it is non-returnees you need to attract in order to get enough people. So, in theory, a plant that treats people well enough that they don't mind coming back would actually have to offer more money to the NON-returnees.
Does that make sense? NO!
How about this --- First, you pay people the same money for doing the same job. If you hire two people who are both capable of covering high-risk work (like S/G's or Refueling) you should pay them the same, PERIOD!! If you have two or three people who are so physically limited, or unreliable, or unskilled, or inexperienced, that they can't do anything but escort laundry and trash, why should they be getting paid the same as the heavy-hitters?
Second, you pay everybody enough to be competitive with all the other jobs they have to choose from. You don't base your pay rates on the tourist attractions in the area, the climate, or the fishing, or "we won't work you as hard here".