Arrhenius equation can NOT be directly used for a non-stationary process kinetics.

You can only apply Arrhenius to the values of reactions rate CONSTANTS, which may be estimated from measurements of the reaction rate at THE SAME reaction conditions (except for T) and at THE SAME conversion extent. This approach is exactly what Ozawa-Flynn-Wall method does.

First of all, you have to measure TPR (or thermal curves of some other process) at several heating rates (e.g. 1, 3, 5 and 10 K/min). So, you have to perform a SERIES of TPR measurements.

Then you can only calculate the APPARENT activation energy of the process.

Three main methods can be used: Kissinger, Friedman and Ozawa-Flynn-Wall.

The first one is based on the shift of the peak maximum position at changing the heating rate and gives you one value of apparent Ea as

ln([heating rate]/[maximum temperature]^2)= - Ea/(R[maximum temperature]).

Kissinger is far from being precise, but is works well even if you have problems with accurate subtraction of baselines.

The two others are more complicated to describe here, you may find OFW method as ASTM E698. http://www.astm.org/Standards/E698.htm. OFW will give you dependence of apparent activation energy on the extent of the process. This is very helpful when you have a two-stage process (for example, when the transport stage of diffusion of cations from the oxide to the surface is followed by the chemical stage of its reduction.

Please note that all these methods were developed for calorimetric thermal analysis.

However, I believe that you can use them for TPR.

Answering you second question - NO. According to my experience (DSC of copper-containing oxides reduction), reduction processes can have very different apparent activation energies and still occur at nearly the same temperature. At that, processes occuring at different temperatures can have similar activation energies.

Run your reaction under two different T conditions and use Arrhenius equation

Arrhenius equation can NOT be directly used for a non-stationary process kinetics.

You can only apply Arrhenius to the values of reactions rate CONSTANTS, which may be estimated from measurements of the reaction rate at THE SAME reaction conditions (except for T) and at THE SAME conversion extent. This approach is exactly what Ozawa-Flynn-Wall method does.

This is for TGA, but I think the basics are the same: H. E. Kissinger Anal. Chem. 29 (1957) 1702

In answer to your second question ... When all else in your measurement system is exactly the same and the overall activation energy of a given process changes, the position of the peak that you measure in TPR for that process will shift. The focus is on OVERALL activation energy, which itself is comprised of a combination of activation energies from each of the molecular steps in the process.

I agree with Khassin's suggestion, using the ASTM E698, so that a referred method can be used for the comparison can be made.

Sure, Alex KHASSIN provided best answers. Est' esche UMY v ROSSII.