Modeling Methodology for Florida Merchant Generation

Ben:

Thomas Suffield has determined, through discussions with our accounting
group, the project cost methodology used for our Genco assets. Unless there
is a compelling reason, we need to conform your Florida model to this model
as (i) it is already an accepted methodology used by the organization, (ii)
it is used as the basis upon which we determine accrual income for our
existing merchant portfolio, and (iii) it will result in more favorable cost
profile in terms of $/kW month than other methodologies.

Please let me know if you have any issues relative to this approach. If not,
please adjust the Florida model and provide me with the results which Mike
and I need to present to Delainey on Friday. It would also be helpful to have
the bid and ask from the desk (5 yr. term) for this presentation. Lastly,
FYI, we have released ICF Kaiser to prepare a forecast specific to the LM6000
configuration (6x and w/wo a 74.9mW steam turbine), which we can then use to
compare to the forecasted cost basis of your model.

Thanks,

Ben