Per our conversation, here is the model that I have for Simon.  The notes 
that I gave you were from some work I did back in September as I was looking 
for volume numbers.  The notes show what drives the final cash flow numbers.

 Briefly, the issue surrounds exactly what is meant by nonmarket and 
noncredit risk.   Ideally, this would be anything that effects NPV of the 
deal.  However, I think that we should limit ourselves to those things that 
effect the time and amount of the EAM volumes.  Even here, such a problem set 
is quite large.  If you glance at the spreadsheet model, you will notice that 
there exists a large number of possible items that effect the EAM volumes.  
If we are to do this rigorously, then it is necessary to tear apart EES's 
business model, and although such an attempt to do so would be noble, the 
scope of such an excercise may be too large.  We need a clear definition of 
EES Operational Risk.

 Briefly, from the EES deals that I have looked at, two things drive their 
profit:  a long term bet that power prices will go down, and that we can 
improve the facility through various enhancements.  Each of these may be 
gleaned from the spreadsheet, as well as the assumptions regarding the 
funding of the facility improvements and so on.  95% of the savings is 
assumed to come from power, and 5% from gas.  (The effeciency gain may not be 
explicitly given).  If we have data that show the realized efficiency gains, 
then it would be simple [in principle] to determine a distribution, and hence 
a distribution of EAM volumes, NPV's, etc.  At this time, I understand that 
RAC has determined some of the realized efficiency gains, but my knowledge is 
quite sketchy.  Jay Hachen may have more info.

 If I get a chance, I will try to see if I can do a "proof of concept" 
excercise, but I have a late January deadline on something else.

 Another point:  even if we are successful in doing this for the spreadsheet 
model, EES has chosen to book things differently.  I do not have a thorough 
understanding of their IT systems, but at least in principle, if we can do 
this for the spreadsheet model, then we can do it in their IT environment 
(data are data are data).  They may book efficiency gains through improvement 
type, such as gains due to compressors, chillers, boilers, etc.  If this is 
indeed the case, then we need to have distributions for each type of 
improvement.   

 If we are successful on the spreadsheet and not successful with their IT 
systems, then the other alternative is to build our own reporting system.  It 
would be similar to a database where the recordsets are replaced by Excel 
workbooks.  It can be constructed in such a way as to enable us to run 
simulations and queries, but this would probably take me about five or six 
weeks.

 Finally, Don Hawkins does Operational Audits for Enron's physical assets.  
He sends out teams to audit our pipelines and strategic assets.  I don't 
think that he does it for EES, but you might want to give him a call anyway.

-Kevin K.

PS:  The EES lunch meeting has been moved to Wednesdays.  Jay Hachen will 
know more.