XTREME INTELLIGENCE INC. / RESEARCH PROTOTYPE 0.1

GridConnect AI

A reproducible view of large-load constraints, mitigation tradeoffs and the limits of a preliminary screen.

Facility demand
Lowest bus voltage
Highest branch loading
Maximum apparent power at either end / RATE_A
Illustrative capital cost
Selected hardware only • synthetic assumptions
01 / Network state

Where the constraint appears

Schematic layout, not geographic. G = generator. Purple ring = facility connection.

Within screenViolation
02 / Voltage profile

Every bus, against the baseline

Filled markers = selected interval; gray markers = public baseline. Shaded area is the chosen voltage band.

03 / Explanation

What this result means

    04 / Constraint evidence

    Failed checks in this interval

    05 / All candidates

    Compare service, constraints and cost

    Worst electrical value over the entire event. Reduced-load and full-load options serve different objectives; no automatic “cheapest” recommendation is made.

    ScenarioFacility MWFull request?Min puMax %ScreenCapex $M*

    *Synthetic capital assumptions. Does not include all project costs or establish economic feasibility. Battery passes apply to this event only.

    06 / Auditable outputs

    Electrical results and assumptions

    Branches — both ends checked
    BranchFrom MVATo MVARATE_A MVALoading %
    Generator dispatch and limits
    BusP MWP min / maxQ MVArQ min / max

    P and Q limits are screened after solving. The program does not enforce Q limits by switching PV buses to PQ; any violated generator limit fails the screen.

    Cost components and exclusions
    Reproduction and model provenance

    Run from the extracted project:

    python -m gridconnect run --config configs/demo.json --out outputs

    This report displays precomputed solver results. Change the JSON configuration and rerun Python to calculate a new study. No electrical equations are approximated in the browser.

    07 / Scope boundaries

    What this prototype does not establish

    Built for Xtreme Intelligence Inc. • Evidence from PYPOWER 5.1.19 AC calculations • Research target: DOE Fall 2026 Topic 14. See the accompanying research brief for funding and competitive evidence.