EXHIBIT 99.376

                  ISSUES AND QUESTIONS WITH CONGESTION PROPOSAL

Potential Issues:

While all details of this proposal may not have to be expressly defined to
initially sell this solution, it must be noted that issues identified by the PX
as minor while in the development of previous projects have had significant
affects on the success of their implementation. It should thus be a priority to
understand, define and resolve all potential issues as soon as possible.

This model relies upon the secondary trading of FTRs to enhance liquidity.
However, it is unlikely that customers who purchased annual FTR contracts will
be willing to wait until the Day-Ahead market to lock-in commitments for them.
This is especially the case in a use-or-lose model such as is proposed for
California. If no FTRs were present in the market, orders will have to be
matched within the same location, as is the case in the current BFM model. The
forward trading of FTRs could potentially enhance their liquidity. [As suggested
by John Flory, forward trading of FTRs/ETCs could be combined with forward
trading of Basis Difference Contracts to enhance liquidity.]

Transmission rights holders could use the ability to recall and modify firm
schedules to greatly exaggerate non-firm market volatility. If firm schedule
modification had too great an effect on non-firm transactions, public confidence
in this market could be undermined. This not necessarily an un-desirable
outcome. [Non-firm energy markets are just that, non-firm. Historically, either
side could recall the transaction at any time, even in real-time. One option of
this proposal is a slight modification, in that trades would become firm at the
close of the hour-ahead market. The other option would treat the NF transactions
as real-time supplemental energy calls in the present market.]

The manner in which RMR pre-dispatch and over-generation are incorporated into
this solution still need to be finalized. [Generally speaking, there does not
appear to be any reason it would not work as today.]

This model does not currently reduce customer uncertainty regarding wheeling out
charges. This has been identified as a market limiting factor and needs to be
minimized. [We should add this as a criteria if folks feel strongly that this
needs to be solved as part of the reform process. We could make it a simple
matter of netting schedules before we send them to the ISO if indeed (subject to
check) the ISO is still charging for wheeling out before it nets schedules. In
any event, it might be a good idea to do this as a precaution anyway, but it
needs thought about how to handle the tags.]

This solution requires the development of a complex algorithm. The PX needs to
ensure that this can, in fact be developed in a manner that retains customer
confidence in the clarity of the PX's prices auction results. [We need to define
transparency. Transparency in the opening auction means that any trader can
self verify that at the posted prices, and for the bids it offered, it could not
improve its profits with a different



                  ISSUES AND QUESTIONS WITH CONGESTION PROPOSAL


set of awards. One exception could be that in the event of a tie (over-lapping
bids, as in the case of over-gen zero bids), the auction might accept part of a
block at the clearing price.]

Questions:

- -        What potential gaming effects are created by the ability to modify firm
         schedules repeatedly until the close of HA? What are the effects on the
         non-firm market? (see above)

- -        How viable is this solution if the expected transmission liquidity does
         not develop? [Or, what if the physical rights model and 100% ATC
         release occurs and the PX does not provide a secondary market?] What if
         the ISO doesn't run the non-firm market? [Or, should the PX be
         proposing to run the non-firm market?]

- -        Are orders into the non-firm market by resource, by portfolio or does
         it matter? (I'm presuming by resource to eliminate additional
         scheduling requirements). [Depending on the congestion model the ISO
         adopts - full network modeling or simplified model - it may have to be
         by resource or it could optionally be by resource.]

- -        Could the PX solution be engineered to allow for linking of hours? [To
         accommodate inter-temporal constraints in the DA auction you mean? We
         had this in the Titanium proposal. It has some implications we should
         discuss.]

- -        Is this the simplest solution that could have been obtained? Are there
         unnecessary complexities that affect ease of implementation / use?

- -        This solution closely resembles the Titanium proposal. What were the
         characteristics of the proposal that led the PX to choose not to
         implement it? How are those issues resolved in this solution?

         [Don't touch the milk-cow. Not a matter of choice now.

         The ISO defended release of only part of the ATC as FTRs. Could happen
         again, but probably not, especially if the PX lines up with the release
         folks.]