Congestion in ERCOT Is Becoming More Time-Dependent

Congestion in ERCOT Is Becoming More Time-Dependent

ERCOT congestion has always been a defining part of the nodal power market, but the way it shows up across the day is changing. Congestion patterns that used to be reasonably approximated using broad Time-of-Use groupings are increasingly time-shaped and dependent on hour-specific system conditions, especially as solar penetration and battery dispatch grow.

This change matters most for solar resources, standalone battery energy storage systems (BESS) and hybrid or co-located projects, because their exposure is inherently shaped by when they produce and when they dispatch. As a result, congestion risk is increasingly something that needs to be evaluated hour by hour rather than treated as a block-average outcome.

Why congestion is becoming more time-shaped

ERCOT now transitions through distinct operating regimes within a single day, and those transitions are being amplified by renewables. Solar introduces a large, predictable injection of energy during daylight hours that changes system flows and alters which transmission constraints bind. Later in the day, as solar output declines, the grid shifts again as thermal resources ramp to serve net load, often producing a different marginal unit and a different congestion regime.

Storage reinforces this time-dependency because batteries concentrate charging and discharging activity into a narrow set of economically meaningful hours. Even when storage is not “causing” congestion, it is highly exposed to it, because nodal price shape and constraint behavior can determine whether spreads are available and whether dispatch value shows up at the expected location.

What this means for CRRs and basis risk

CRRs remain the primary tool for hedging congestion exposure in ERCOT, but time-dependent congestion reduces the effectiveness of treating Time-of-Use products as internally uniform. When multiple congestion regimes exist inside the same hour set, a flat hedge can miss the hours that matter most, even if it looks reasonable on an average basis.

This dynamic is most visible for solar because production is concentrated in the same midday window where congestion behavior can differ from the rest of the on-peak block. It is equally important for batteries and hybrid resources because value is created through intraday spreads, and congestion can materially shift the best charging and discharging windows. For co-located solar plus storage, this becomes a combined problem: the solar component concentrates exposure into a specific part of the day, while the storage component is explicitly trying to monetize the transition periods where congestion and price formation can shift quickly.

To illustrate how this can show up in practice, consider a solar asset that is managed under a standard Time-of-Use hedge. For example, a CRR position that is structured around a broad on-peak definition effectively settles against average congestion across the full sixteen-hour on-peak block. The asset’s realized exposure, however, is not evenly distributed across those hours. Solar production is concentrated into a narrower set of midday hours, meaning the effective congestion exposure experienced by the asset is closer to a generation-weighted average than a flat block average.

Hourly P50 generation for an anonymized ERCOT solar asset under management (200 MW), averaged across the year and shown by Hour Ending. This profile is used as the weighting series for the production-weighted congestion example. 

In a simple hub-based example using ERCOT path HB_WEST (source) → HB_NORTH (sink), hourly day-ahead congestion was represented as the difference in the day-ahead congestion component between the two hubs (DACONG at HB_NORTH minus DACONG at HB_WEST). Using one full year of hourly data and an hourly solar production profile as a weighting series, the on-peak block average congestion across the full on-peak hour set was $0.81/MWh, while the solar production-weighted congestion over the same period was $2.73/MWh. This implies a weighting effect of +$1.92/MWh between a flat block-average view and a generation-weighted exposure view.

Day-ahead congestion is shown as the hourly differential in the estimated congestion component between hubs (DACONG at HB_NORTH minus DACONG at HB_WEST), averaged by Hour Ending. Using the solar P50 profile of the 200-MW solar resource (blue line) illustrates that solar exposure is concentrated in a narrow midday window, where congestion outcomes can differ materially from the broader on-peak block average.

Importantly, the +$1.92/MWh is not the standalone value of a CRR position. It represents the gap between two ways of measuring effective congestion exposure: a flat on-peak average that weights each on-peak hour equally, and a production-weighted metric that assigns greater weight to the hours when the asset actually produces. Applied to an annual P50 production volume of 421,236 MWh, this difference corresponds to approximately $810,863 per year of effective congestion exposure that would be understated if congestion is evaluated using only a block-average interpretation.

How market participants should adjust

The practical implication is that congestion exposure and hedge effectiveness need to be evaluated using the same weighting the asset experiences. For solar, that means analyzing generation-weighted basis and settlement outcomes instead of relying on block-level averages. For batteries and hybrids, it means incorporating hour-specific congestion regimes into expected value, dispatch strategy, and hedge design, rather than assuming the same location relationship holds across an entire Time-of-Use strip.

For co-located solar plus storage projects, the need for time-aware evaluation is even more pronounced. Solar output is concentrated into a predictable midday window, while storage economics are often driven by morning and evening transition periods. If those windows sit in different congestion regimes, a single block-level hedge or a single block-level performance lens can appear reasonable on average while still failing to align with where the project actually realizes value and where it experiences risk.

At a minimum, participants evaluating congestion risk and CRR hedge effectiveness should segment congestion outcomes within existing on-peak definitions into sub-periods that reflect how resources actually operate. Even without any changes to market products, this type of segmentation helps clarify whether congestion is being realized primarily during solar hours, during non-solar peak hours, or during transition periods where storage tends to be most active.

Closing thought

ERCOT congestion is becoming more differentiated by time of day because ERCOT system conditions are becoming more differentiated by time of day. Solar, BESS, and hybrid or co-located resources are the most exposed to this shift, which is why they will feel it first. The participants that adapt fastest will be the ones who treat congestion as an hourly risk profile rather than a single block-level assumption.