---------------------- Forwarded by Vince J Kaminski/HOU/ECT on 12/22/99 
04:34 PM ---------------------------


Clayton Vernon@ENRON
12/22/99 10:40 AM
To: Vince J Kaminski/HOU/ECT@ECT
cc: Vasant Shanbhogue/HOU/ECT@ECT 
Subject: Re: working gas price model  

Vince-

I have a simplified version of Brad's model in mind.

The "no arbitrage" condition equates trading margins across the country. 
Costs of transmission rise with congestion on the network. Wellhead supply is 
almost completely price- elastic, while burner-tip demand is almost 
completely price inelastic. Storage is rationalized as a perpetual call 
option.

The least time-variant parameters are the costs of injecting and withdrawing 
gas from storage to the pipeline, followed by the costs of delivering gas 
from the wellhead to the pipeline. The intermediate-variant parameters are 
the capacity-dependent costs paid to the pipeline (above shrinkage) for 
transmission. The most time-variant parameters are the trading margins and 
the valuations of the storage option.

There are 8 parameters to be estimated at each major node of the betwork. 
They are identifiable in either of two straightforward ways: using a short 
time series of the last 3 days prices based on the assumed variability 
mentioned above, or point-estimates ("calibrations") using only today's data 
based on a node-based model of competition between pipelines where pipes with 
the same region of origination, albeit markedly different terminus, price 
versus capacity similarly, "competing" for outflows.

I will write this up for you in Scientific Word and present it to you at your 
earliest convenience.

Clayton