Outstanding issues with co2_adsorption_desorption example

[dallan-keylogic]

1. The time horizon reduction step as presently written does nothing for the adsorptions simulation because it occurs after time_set is created. It probably goes into effect for the desorption step.
2. A property is currently being created in a way that IDAES does not like, causing a depreciation warning. This is an easy fix, because an example about how to include it properly is right below it.
3. There are presently a lot of badly scaled equations
4. There are unit errors in the equations. Unit errors in the time discretization equations are a known issue, but there are also errors in reaction rate equations, etc.
5. The final time used to index Pyomo elements comes from the PETSc time variable. Theoretically, these are supposed to be the same, but whatever tolerance PETSc uses is weaker than the one Pyomo uses, so you get a KeyError when the code tries to use an index of 101810.60189669262 when Pyomo is expecting one of 101810.
6. The heat_computation calculations need an overhaul. The spatial integral used to get the instantaneous heat input to the adsorption column is declared after the ContinuousSet is discretized, which is undefined behavior. The time integral used to get the total heat input over all time is done by creating a second time set which is then discretized, which is a very awkward procedure. @blnicho recommends not using Pyomo Integral objects anyway. I recommend embedding the spatial integral in the FixedBed1D object with a manual quadrature rule as a DerivativeVar of the total heat added, then allowing PETSc to integrate it while it integrates everything else.