18 Oct 2026
MISO: $396 NSP Congestion Spread Prints With No Named Constraint
The widest real-time spread across MISO in the 21:05 UTC interval on Saturday, 2026-10-17, was $396.42/MWh, and both ends of it sat inside Northern States Power. NSP.UOFMGEN1 cleared at $169.18/MWh; NSP.ALDRIHERC cleared at $-227.24/MWh. That is 16:05 Central on a Saturday afternoon, a single five-minute interval, and it is the only price observation in the record: no 2026-10-18 LMP has printed in the source data yet.
The decomposition leaves no room for an energy story. The MISO system energy component was $9.10/MWh at every listed node, losses ran between $-0.71 and $-0.77/MWh, and everything else was congestion. UOFMGEN1 carried $160.82/MWh of congestion, and NSP.HENNIPIN1 printed an identical $169.18/MWh with the same $160.82 component, which marks the two as electrically coincident behind one constraint. The gradient steps down from there: NSP.ELLIOTPK1 at $148.52/MWh on $140.17 of congestion, NSP.UOFMGEN2 at $131.70/MWh on $123.35, and NSP.MERPK1 at $69.35/MWh on $60.99. On the other side, ALDRIHERC's $-235.57/MWh congestion component swamps the energy term entirely, and the four OTP.RUGBY1 nodes (MRES, UMMP, UPLS, IBR) all cleared at $-73.82/MWh on $-82.21 of congestion, a second, shallower negative pocket. The 7-day MISO real-time binding-constraint feed logged at least 7,969 shadow-price events and the estimated outage feed at least 60, with both counts possibly truncated, but the sampled constraints (an AMI34034 flowgate, the Overton 345/161 transformer, Tupelo Tap 138 kV, Lime Creek-Barton 161 kV) sit outside the NSP and OTP footprints, and the sampled outage rows carry no region label. The element binding behind the $160.82 and $-235.57 components is unidentified, and so is whether ALDRIHERC's print reflects curtailed output, a unit outage, or a line outage.
For the angle to cash, the spread has to outlive a single Saturday interval, and the first clean test is the same clock position a day later. If NSP.UOFMGEN1's congestion component still exceeds $160.82/MWh in the 16:05 Central interval today, 2026-10-18, the constraint is persistent rather than a one-off and the UOFMGEN1/HENNIPIN1 pair should be priced as one leg. If UOFMGEN1's LMP instead sits at the $9.10/MWh system energy component at 09:00 Central, the congestion has cleared and the basis premium goes with it. On the sink side, measure NSP.ALDRIHERC's congestion component against its own $-235.57/MWh Saturday print at 16:05 Central today: a reading at or below that level means the depressant held, while an LMP back at the $9.10/MWh energy component confirms a one-interval spike. Separately, check whether the OTP.RUGBY1 cluster is still pinned at $-73.82/MWh at the same time; two negative pockets rather than one widens the set of candidate elements. Day-ahead LMPs for these nodes are absent, so the DA/RT basis for the interval cannot be computed yet.
> A $396 nodal spread on a $9.10 energy term is a transmission price, and until MISO names the element behind it, the NSP basis is pure congestion risk.
Not investment advice. For informational purposes only.
Grid Alpha — daily briefs (https://feed.gridalpha.us/?utm_source=podcast&utm_medium=shownotes) · live dashboards, all nine markets (https://app.gridalpha.us/?utm_source=podcast&utm_medium=shownotes) · public call record (https://feed.gridalpha.us/?utm_source=podcast&utm_medium=shownotes#call-record)
