Sioshansi F.P. Smart Grid: Integrating Renewable, Distributed & Efficient Energy
Подождите немного. Документ загружается.
directly the bill impacts of LMP-based rates, it was a
challenge to develop market rules for them to participate in
the Economic Load Response Program. The solution required
their load reduction capabilities to be “dispatchable” by PJM
in real time. Following dispatch of the DR, PJM assumed the
risk of falling LMPs, which ensured that the end-user would
realize savings at least equal to the savings calculated on the
basis of the “strike price” it provided to PJM. If, for example,
PJM dispatched a load reduction with a “strike price” of
$100/MWh for four hours and the LMP fell from $150/MWh
in hour one to $50/MWh in hours two, three, and four, then
the PJM market would make the participant whole for the
anticipated savings of $400.
Market Results
The trends in activity in terms of MWh reductions and
settlements paid out on a monthly basis are shown in Figure
17.8. As can be seen in Figure 17.8, throughout 2006 and
2007 MWh reductions increased and so did the dollar value of
settlements. Activity and settlements remained fairly steady
through the first seven months of 2008 and have fallen off
precipitously since then.
841
Figure 17.8
DR activity in PJM's energy markets 2006–2011.
Source: PJM
The increase in activity coincides with a period of increasing
energy prices overall in PJM during the 2006–2008 period.
Beginning in September 2008, which also coincides with the
recognition that the economy was falling into a deep
recession with the financial crisis, PJM energy prices also fell
and have remained low relative to levels seen during the
2006–2008 period. Figure 17.9 shows the monthly evolution
of PJM load-weighted average LMP from 2006 to 2011.
842
Figure 17.9
PJM load-weighted average LMP 2006–2011.
Source: PJM
Moreover, some activity prior to 2008 was likely driven by
the incentive payments available to DR in the Energy Market
for load reductions occurring at prices above $75/MWh. This
incentive was allowed to expire at the end of 2007 per the
terms of the PJM Tariff and Operating Agreement.
24
24
See also the denial of a complaint that sought to extend the incentive
payments: 121 FERC 61,315.
Economic Load Response participation is shown in Table
17.3. It is important to note when analyzing the payment data
that from 2002 through October of 2007 the incentive
payment was of the full LMP rather than LMP‒G&T for load
843
reductions in hours when the applicable LMP equaled or
exceeded $75/MWh. Payment of the incentive required PJM
to collect the difference between the LMP and G&T from the
end-use customer's LSE, and the G&T portion of the payment
from all of the LSEs in the zone on a weighted basis. The
allocation of revenue among Economic Load Response
activity in the day-ahead, real-time dispatchable, and
real-time self scheduled options has ranged on a percentage
basis in the low single digits except for the real-time self
scheduled option, which consistently exceeds 90%.
Table 17.3
Annual Revenues Paid to DR in the PJM Energy Market
2002 2003 2004 2005 2006 2007 2008 2009 20
$801,119 $833,530 $1,917,202 $13,042,865 $18,334,295 $66,987,730 $27,759,624 1,274,675 $3,08
Generally, the total revenues paid to CSPs between 2002 and
2009 reflect not only the impact of the incentive payment
described above but also the growth in the number of
participating CSPs and of registered and active end-use sites
shown in Figure 17.10. Even with the decrease in energy
prices and therefore more limited opportunities for load
reductions, the level of registrants has remained fairly
constant.
25
25
The dips in registration seen in Figure 17.10 are due to registrations
expiring at the end of the delivery year (May 31) and DR participants
re-registering just after June 1 of each year.
844
Figure 17.10
MWs of DR registered to participate in PJM's energy markets.
Source: PJM
From the perspective of an individual end-use site, the
revenue opportunities for DR participating in Economic Load
Response during the high-price years of 2007 and 2008 can
be seen in the following example. During the 2007 calendar
year a 1-MW DR with a retail rate
26
of $75/MWh and a
willingness to reduce demand by 1-MW each time LMP was
at or above $100/MWh would have collected $44,443 in
revenue over and above the savings from avoiding the retail
rate during the Load Reduction Events. In 2008 when prices
were higher, especially during the first seven months of the
year, the same DR acting in the same way would have
collected $86,820 in revenue over and above the savings from
avoiding paying the retail rate during the Load Reduction
Events.
26
845
For the purposes of settlement in PJM, the retail rate is the generation
and transmission portion of the retail rate also known as G&T.
Emergency DR Participation in the Energy Market
DR registered as either Emergency Energy Only or Full
Emergency Load Response receives an energy payment for
reductions when called by PJM Operations to respond to an
emergency event. Response by Emergency Energy Only DR
is voluntary, while response by Full Emergency Load
Response is mandatory. Full Emergency Load Response
resources also receive a capacity payment for their
willingness to reduce load at the direction of PJM Operations
a maximum of 10 times during the summer period for up to
six consecutive hours. The load reduction quantity used to
calculate the energy payment associated with these load
reductions is measured by comparing the metered load during
the full hour before the hour in which PJM Operations calls
for Emergency DR with the metered load during each hour of
the load reduction event. FERC Order 719 mandated the
removal of imbalance penalties for LSEs whose loads fall
below day-ahead schedules when system usage is at its peak
and required reserves are short, in order to remove any
disincentive to responding to high prices when it helps
maintain reliable operations.
27
CSPs cannot easily anticipate
precisely when PJM Operations will call for Emergency DR
so that the usage level at the site during the full hour before
the hour in which the DR is called can be the CBL used for
the energy settlement calculation.
27
See Wholesale Competition in Regions with Organized Electric Markets,
Order No. 719, FERC Stats. & Regs. ¶31,281 (2008) and PJM
846
Interconnection, L.L.C., Initial Filing, Docket No. ER09-701-000
(February 9, 2009).
Order 745, Demand Response Compensation in
Organized Wholesale Energy Markets, Issued 3/15/11
The FERC's recent Order 745 imposes short-term and longer
term Economic Load Response compensation requirements
on organized wholesale electricity markets. The FERC has
determined that DR responding to Energy Market prices
should be paid the Locational Marginal Price (LMP) for
achieved load reductions but only (1) when the RTO or ISO
can use DR to balance supply and demand as an alternative to
generation and (2) when dispatch of DR results in benefits,
measured by the savings to market buyers from a lower LMP,
that exceed the cost of total payments to the CSPs. The Order
further requires the RTOs/ISOs to develop a methodology for
calculating in advance for each month the LMP threshold at
and above which the benefit exceeds the cost. Finally, the
RTOs/ISOs must study the feasibility of implementing a tool
that determines dynamically the LMP threshold.
28
28
Demand Response Compensation in Organized Markets, 134 FERC
¶61,187 (March 15, 2011).
If Order 745, as PJM believes, directs organized wholesale
markets to pay full LMP when DR responds to PJM dispatch
directions in the same way as generation, then CSPs will have
fewer options for participating in the Energy Market and will
need to alter their current operations significantly. Currently
as explained above the CSP can offer the load reduction
capability (1) in the Day-Ahead Energy Market or (2) for
847
dispatch by PJM in real time; or it can (3) self-schedule the
load reduction through the notification process. Historically,
CSPs have used the self-schedule option approximately 98%
of the time. The removal of the self-schedule option from the
market rules will require CSPs to adjust how they interact
with end-users whom they have registered as Economic Load
Response.
Resolving Early Issues with CBL Estimation
Estimating what the end-use site would have used but for the
demand response activity, known as the CBL, is the single
most difficult challenge for DR participation in the wholesale
Energy Market. Reviews of some settlements led to questions
about whether the Economic Load Response activity was real
or “happenstance,” that is, simply the result of running meter
data through the “high 5 of 10” Customer Baseline (CBL)
calculation in the market rules. Concerns also arose about
stale CBLs resulted from the standard “high 5 of 10” CBL
calculation and about end-users whose load varied
significantly from day to day. The “high 5 of 10” rule
calculated the CBL by averaging the five highest hours
corresponding to the load reduction event of the most recent
10 qualifying days.
Concerns were raised in late 2006 about the behavior of some
participants in Economic Load Response as expressed in the
following question, “Did the submitted settlements (support
for an Energy Market payment) reflect a change in the
end-user's normal usage pattern in response to hourly Energy
Market prices or did gaps in the market rules permit payment
of Economic Load Response settlements inappropriately?”
Such a gap in the market rules would have threatened the
848
integrity of Economic Load Response by DR in the Energy
Market.
PJM stakeholders and staff worked together to develop tariff
and manual changes designed to protect the integrity of
Economic Load Response in the Energy Market, to permit
aggregation of qualified end-use sites in order to meet the
required 100-kw threshold, and to enable CSPs to submit up
to 10 price, quantity pairs for DR mirroring precisely the
offers of generators.
29
This work involved a comprehensive
evaluation of a number of CBL methodologies. The criteria
for the evaluation included accuracy, bias, and simplicity. The
Commission issued an order approving the proposed tariff
revisions in mid-2008.
30
29
PJM Interconnection, L.L.C., Initial Filing, Docket No. ER08-824-000
(April 14, 2008).
30
PJM Interconnection, L.L.C., 123 FERC ¶61,257 (2008).
PJM staff subsequently developed a set of objective tools for
uniform enforcement of the revised tariff and manuals. PJM
flagged end-use sites when CSPs submitted Economic Load
Response settlements for less than $5 or more frequently than
70% of the days on a rolling 30-day basis. CSPs with flagged
end-use sites must provide PJM with a copy of the contract
for curtailment services, a detailed description of the how the
end-use site reduces load during an event, and documentation
that load reduction activity took place in response to hourly
wholesale energy prices. These tools have been effective in
aligning CSP activity with PJM expectations and in
discouraging the submission of “happenstance” settlements.
849
DR Participation in Ancillary Service Markets
Synchronized Reserve Market
Synchronized Reserve (SR) service is necessary to serve load
immediately in the event of a system contingency or
unexpected need for more power immediately (“Synchronized
Reserve Event”). In PJM SR is broken up into Tier 1 SR,
which is provided by units that are on-line, following
economic dispatch, and capable of increasing output within
10 minutes of a call for SR, and Tier 2 SR, which is extra SR
capacity committed in excess of Tier 1 capacity to meet SR
requirements. The SR Market is the mechanism by which SR
is committed.
In the same FERC Order that made the Economic and
Emergency Load Response Programs in the Energy Market
permanent, the rules were put in place for DR to participate in
the SR Market.
31
In order to participate in the SR Market,
DR must be capable of dependably providing a response
within 10 minutes of a Synchronized Reserve Event call and
must have appropriate metering infrastructure that provides
data at no less than a one-minute scan rate surrounding a call
for SR. DR must also be able to maintain its load reduction in
response to a Synchronized Reserve Event for up to 30
minutes. Cleared DR that responds to a synchronized reserve
event must provide one-minute meter data from 10 minutes
before the event to 10 minutes after the event ends. PJM uses
the meter data to assess the performance of the resource by
comparing the cleared MW with the difference in metering
readings between the time that PJM notifies the CSP of the
event (Minute 1) and the beginning of the event (Minute 10)
as well as for all of the minutes (up to 30) of the event. PJM
850