scheije@selectenergy.com writes to the NYISO_TECH_EXCHANGE Discussion List:

I apologize for the larger circulation of this email than to just the S&P
WG.

The MC last week urged a rapid consideration of the interactions between
BME and SCD and their related prices from the hot, high load days of July
24 and 25.

To pursue this there is additional information needed from the ISO. Below
is a preliminary list I compiled that would be useful in assessing the
situation.

Data should be provided for relevant hours i.e. those with high BME or RT
prices or both. Particularly for those hours with binding reserve
constraints in HA or RT.

1 RT reserve levels in each hour (it is recognized that the value
fluctuates but the the point is to get an idea of apporximately how short
the ISO was of reserves, if at all)

2 Similar information should be obtained for our neighbors; PJM, NE and
IMO. Also, could the ISO describe what the others do hour ahead to prepare
for the next operating hour under circumstances such as we encountered
those two days? NY has the unfortunate position of producing prices related
to decisions for which it pays otherwise - what were our neighbors doing?

3 An accounting, by hour, of reserves carried on exports out of NY

4 Identification of any non-ICAP based exports that may have been counted
for reserve; this shouldn't happen as it puts non-ICAP providers in the
position of being recalled...

5 BME reserve deficiencies by hour; were the reserve levels always met,
albeit at extreme prices, or was the ISO unable to cover the requirements
in advance?

6 Based on Mike Calimano's report, there were "a couple hundred Mws" of
operating units not seen by BME; this situation needs to be (1) confirmed
and (2) explained and corrected

The following are sort of 2nd tier questions:

1 If the ISO was short of reserves in RT why weren't the EDRP resources,
who were notified, not called into play to aid in curing the deficiency?
There should have been no economic judgement call by operators (if the
first two hours were at $500 that was likely cheaper than what BME was
taking in if the BME prices applied to settlement under ECA B). BME made no
judgement call when its prices spiked due to the 30 M NSR scarcity and took
available imports.

2 What would have been the System State(s) if the reserve margins had been
held in RT (assuming ISO was short in RT)? Would this have forced a Major
Emergency and subsequent voltage reductions, public appeals etc? Had this
been done would have the EDRP Mws been activated?

I would also like to offer the following as potential changes to be
considered. They are based on the limited discussions that have taken place
so far.

1 Include exports that are being counted as reserve in BME as it evaluates
its reserve position. Since BME is deciding which transactions will be
scheduled, operators would need to indicate which transactions and how many
Mws could be counted in RT as reserves. The implications for the solution
algorithm are unclear - can BME recognize the additional reserves thus
carried? Would it result in fewer imports and lower BME prices? What if the
reserve requirements used as input to BME were manipulated downward for
reserves on exports? What happens if BME did not schedule the export(s)?

2 Subject to the ISO providing detailed RT reserve info, it seems that the
reserves should be carried in RT and that SCD should not be allowed to
dispatch with units that were not available to BME (thus resulting in the
large price disparities). If the ISO had adequate reserves in RT, SCD
should not be allowed to produce RT price reductions relative to what BME
is doing. However, having said this, if there are no energy resources to
maintain reserve levels AND the ISO needs to maintain its regulation room,
the next thing to go is load. To my long understanding, we do not shed load
to avoid shedding load (at least not to maintain reserves...)

Perhaps if BME has some adjustments and the missing Mws are accounted for,
little needs to be done in RT. (This does not mean that SCD handles GTs
correctly - that is still a commit problem it is incapable of resolving...)

3 At one point, we had approved use of BME prices to settle all HA
transactions. This approach was scuttled by the ISO when they discovered
that there were persistent, unidirectional price differences between the
proxy busses and their connected in-NY zones. We then retreated to
automating what is done under ECA B; substituting BME prices only under
constrained ramp or ATC conditions.

The ISO should review this again. Does the price difference remain,
inviting gaming? If so, is there a way for the MMU to watch for or prevent
inappropriate opportunistic behavior? A full, partial 3rd settlement would
lessen the sting of HA prices that are out of synch with RT.

4 The ISO should review the multipliers in the LP solution that insure that
energy, then regulation and then the 3 types of reserve are maintained.
They should be no more than required for a proper solution outcome. Prices
at $65000 are not necessary and misleading (while mathematically
correct...). Also, if we are to consider settling transactions at BME
prices, such an outcome would result in considerable opposition...

5 Another non-controversial change would be to extend the DADRP
participants to hour ahead evaluation. In the instant circumstance, it is
likely that BME would have chosen some of those loads opposed to imports
(there was no congestion so some of those western loads might have looked
pretty good...). Since there is the issue that the DADRP is subsidized, I
would propose that, if extended to BME, that the DA bids be used for BME
evaluation. If nothing else look at it as sending money to companies in
NYS.

Jim