While the Midwest ISO’s (MISO) markets continued to perform competitively and efficiently in 2011, Potomac Economics made several transmission-related recommendations in its 2011 state of the market report, including that MISO develop provisions that allow non-dispatchable demand response (DR) – including interruptible load and behind-the-meter generation (BTMG) – to set energy prices in the real-time market.
As the capacity surplus falls in MISO, the system’s peak needs will increasingly be satisfied by interruptible load, BTMG or other forms of DR.
If these resources cannot set prices in the real-time market, MISO will be understating energy prices increasingly during peak demand conditions, Potomac Economics, MISO’s independent market monitor, added, noting that prices in these hours play a key role in sending efficient economic signals to maintain adequate supply resources and to develop additional demand-response capability.
Therefore, allowing demand response to set real-time energy prices will improve incentives to schedule imports and exports, to schedule load in the day-ahead market – and reduce revenue sufficiency guarantee costs – and to invest in resources needed to maintain adequate supplies in MISO.
Potomac Economics also said that this recommendation was proposed in 2008 and that MISO has worked to address it by allowing emergency demand response (EDR) to set prices through the extended locational marginal price (ELMP) initiative, which is scheduled to go into testing in early 2014.
While the progress in potentially allowing EDR resources to set prices as part of the ELMP initiative has been substantial, it is important to address load-modifying resources (LMR) and BTMG that will be deployed first in an emergency. This may be done by establishing a default curtailment cost for each class, or by compelling those resources to participate in the EDR program. The latter approach, Potomac Economics added, has the advantage of providing MISO more direct access to these classes of DR capability, and perhaps an improved capacity to verify their ability to curtail load when needed.
Another recommendation called for MISO to discontinue the constraint relaxation algorithm for market-to-market (M2M) constraints and set LMPs based on a transmission constraint’s marginal value limit (MVL) when the constraint is unmanageable.
Potomac Economics noted that the constraint relaxation algorithm artificially reduces the real-time congestion that is reflected in MISO’s LMPs when a constraint is violated – for instance, when the real-time market cannot reduce the flow to less than the limit in the interval.
Pricing such congestion accurately is key as it will facilitate day-ahead market outcomes that cause generation to be committed efficiently to manage the congestion. It also provides the necessary economic signals to guide investment in new resources and transmission, Potomac Economics added, noting that in 2011, the constraint relaxation algorithm reduced apparent real-time congestion by one-third, or $620m.
The recommendation was proposed in 2005, and MISO discontinued use of this algorithm for all non-M2M constraints on Feb. 1, 2012. It continues to use the methodology on M2M constraints because PJM Interconnection disagrees with MISO discontinuing its use on those constraints.
“We understand that MISO will continue its dialogue with PJM to determine whether there is hope of achieving a consensus on this change,” Potomac Economics said. “If not, other avenues should be explored, since the use of this methodology is not specified or otherwise required under the [joint operating agreement] with PJM.”
Potomac Economics also recommended that MISO consider implementing a graduated MVL – for instance, a transmission demand curve – for transmission constraints.
Transmission constraints are frequently violated in small quantities or for brief periods of time because the power flows over MISO’s constraints are affected by external factors that can cause them to change unexpectedly. To the extent this causes relatively small violations, it may not substantially affect reliability. If this is true, Potomac Economics added, pricing small violations at the full MVL may not be efficient, and this can be alleviated by replacing the single MVL with a graduated demand curve.
Going forward, MISO should complete its evaluation of this recommendation’s costs. Since 71% of the congestion value that happened on the MISO network in 2011 was on constraints that were in violation, introducing transmission demand curves could lower congestion-related costs. If such curves reflect the reliability implications of violating the constraints by small amounts, it will lead to more efficient real-time pricing. “It is important that the transmission demand curves reflect the system’s reliability needs and that MISO operates the system consistent with the demand curves,” Potomac Economics added.
MISO should eliminate ‘deadband’
Among other things, Potomac Economics called on MISO to eliminate the transmission constraint deadband, which aimed at reducing price and generator dispatch volatility by helping ensure that once constraints were binding, they continued to do so.
However, Potomac Economics said it believes the deadband is increasing volatility for it contributes to unmanageable congestion that often results in sharp LMP changes. Furthermore, it inefficiently reduces the use of the transmission system by binding constraints at levels less than their physical capability. “We are unaware of any other RTO that currently employs a transmission deadband,” Potomac Economics added.
Potomac Economics’ evaluation of the unmanageable congestion in MISO showed that 30% of the value of constraint violations happened when the transmission deadband alone caused a constraint to appear to be violated – for instance, when the flow was less than the original transmission limit.
The deadband accounted for about $140m in unpriced congestion and 19% of all congestion value in MISO during 2011, Potomac Economics added, noting that while eliminating the deadband would not cause the congestion to fall to zero, it would be significantly less.