Cluster Studies: Your Upgrade Bill Depends on Your Neighbors

Cluster studies and the SPP ATSS

Many grid operators study transmission and interconnection requests together, as a group or cluster. The upgrades, costs and in-service dates each request is assigned depend on which other requests enter the same study. This post uses SPP’s ATSS as the worked example; the same cluster analysis applies to other group studies.

In Southwest Power Pool, resources that need long-term firm transmission service may be studied through the Aggregate Transmission Service Study, or ATSS. The ATSS determines whether the requested service can be granted and identifies required network upgrades, while capacity accreditation remains a separate calculation.

SPP opens two ATSS cycles each year and studies the requests in each cycle as a cluster. Assigned upgrades can change both cost and timing: a major obligation can make a resource uneconomic, and a late in-service date can delay when firm service is available to serve a data center or other large load.

Firma replicates the ATSS study exactly for each scenario

For the large-load use case described here, each scenario represents a generation-to-load transfer from named resources to a receiving footprint. FirmaOS solves the applicable seasonal planning cases, tests network performance under normal and single-contingency conditions, and identifies overloaded facilities attributable to each request. The model reflects SPP’s published framework and implementation details Firma Power has refined through direct work with SPP’s engineering team.

At this scale, FirmaOS uses two proprietary systems: a fast, parallelized DC screening engine and a cloud-deployed AC power-flow solver. They allow Firma Power to evaluate large scenario sets while preserving the network detail required for decision-grade results.

One modeled cycle answers a conditional question: what upgrades, cost allocations, and need dates would result if this exact group of requests entered together? It is one possible outcome, not an expected value.

Model a distribution, not a point estimate

The final cohort is unknown when a resource is selected. FirmaOS therefore generates tens of thousands of realistic cluster compositions using historical ATSS submissions, projected data-center buildout, and proprietary generator-selection logic that reflects which assets are likely to pair with transmission requests.

Every composition runs through the same transmission and cost-allocation framework. The result for each resource is a distribution of estimated allocated upgrade cost—not a single estimate, showing where outcomes concentrate and how far the tail can extend. These distributions are decision-support ranges, not forecasts of the exact requests SPP will receive in a future cycle.

Firma uses the distribution to build a derisked portfolio

FirmaOS uses these distributions to identify resources with attractive cost and upgrade-risk profiles, including assets that require a full ATSS. Those candidates remain eligible for Portfolio Designer; resources with excessive tail exposure or schedule risk can be screened out before capital is committed.

That confidence expands the resource catchment beyond only the simplest study paths. Firma Power can consider projects with earlier CODs, stronger permitting or development progress, and better economics—then select which requests to pursue, manage study gates and financial commitments, and feed formal results back into the portfolio. The outcome is a broader, better-underwritten supply pool and a faster path to power.