Power Meter/Sensor uncertainty budget..

[moosebiscuit]

Good morning everybody. I've hit a wall! 

I cannot for the life of me figure out how to quantify the uncertainty for power measurements utilizing an Agilent E4419B and 8482A Power sensor. I've found some comparable scope's of accreditation that give me a reference for what values I should be getting close to. I have attached the part I'm referencing to this post.  With this equipment I don't really know where to start. I've been using a nice range-calculator/budget that's useful for more basic utilities for one standard but I can't seem to figure out how to apply it here. This is what I have so far.

- Past calibration certificates for both the 8482A and the Meter, both accredited calibrations with associated MU's.
- According to the power meter's specifications, the Zero Set accuracy is +- 50 nW.
- Power output is 1.00 mW (0.0 dBm) factory set to +-0.4% traceable to NPL
- Power reference accuracy for two years is +-0.5% with my lab environmentals
- Resolution of the power meter is .0001
- SWR is 1.06 Maximum
- With use of 8482A sensor, Zero drift is < +-10nW and measurement noise <110 nW

Supplemental Data--

From the Power Meter certificates I have power measurements expressed in Watts.
From the Power Sensor certificates I have Voltage Reflection Coefficient tests displayed like so:

The PHASE column is the measured PHASE(Degree).
The MEASURED column is the measured MAGNITUDE.
The MAXIMUM column is the specification of MAGNITUDE.
The UNCERT column (if it is present) is the uncertainty for MAGNITUDE.

As well as calibration factor measurements for the various frequencies, the factor itself, Cal fac uncertainty displayed in %, Refl Coef magnitude and refl coef phase along with this note on the bottom:

The CAL FACTOR UNCERTAINTY (if it is present)
is stated as an ABSOLUTE uncertainty.
The RELATIVE uncertainty can be derived from
the ABSOLUTE uncertainty by the following equation:
RELATIVE CAL FACTOR UNCERTAINTY =
100% * (ABSOLUTE CAL FACTOR UNCERTAINTY)/(CAL FACTOR)

My questions..

1.) Which of this information will need to be applied to an uncertainty budget? How do I quantify everything in the end for use in my CMC?
2.) For repeatability testing, since I am quantifying one number for a range (100 kHz to 4.2 GHz, the range of the sensor) how would I go about that exactly? Tips?
3.) What am I missing?

Thank you kindly.

[ap]

Keysight makes available power sensor calculators. You could use this as a starting point.
Just search for 'keysight power sensor uncertainty calculator'