Simulation Output

All per-scenario simulation results use Hive partitioning: one data.parquet file per scenario stored in a scenario_id=NNNN/ subdirectory. See Hive Partitioning for how to read these files.

Node Axis and Policy-Graph Outputs

Every simulation/ entity partition below shares a leading (scenario_id, stage_id, node_id) axis: scenario_id is the scenario id, stage_id the declared stage id, and node_id the id of the visited policy-graph node for that (scenario_id, stage_id) pair. On a plain stage chain (the default: a policy_graph that declares no nodes) the nodes are numbered by 0-based stage position, so node_id is that position; it equals stage_id only when the declared stage ids are 0, 1, 2, … (declared stage ids 10, 20, 30 give node_id 0, 1, 2). On a policy graph that declares nodes, node_id is the node’s declared id and identifies which branch a scenario actually visited at that stage, and joins back to the node’s stage_id/pool_id via the nodes of the policy checkpoint’s manifest (see policy/manifest.bin).

Methodology: Policy Graphs

Two run-level, unpartitioned files complete this axis:

`simulation/paths.parquet`

The per-scenario node-path trace: exactly the (scenario_id, stage_id, node_id) axis prefix, one row per stage visited by each scenario. 3 columns, all non-null. Joins to any entity file on (scenario_id, stage_id).

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based).
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).

`simulation/scenario_summary.parquet`

Run-level, per-scenario summary: one row per scenario. 3 columns.

Methodology: Upper Bound Evaluation · Discount Rate Formulation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based).
`probability`	Float64	Yes	—	Per-scenario leaf-path weight under a declared census. `null` under sampled scenario selection.
`discounted_immediate_cost`	Float64	No	USD	Present value at the first stage of the scenario’s immediate cost: the sum over the visited stages of `discount_factor` × `immediate_cost`; when a boundary policy is loaded, the last stage’s term uses `total_cost` in place of `immediate_cost`.

`simulation/costs/`

Per-stage cost breakdown. One row per stage. 29 columns.

Methodology: LP Formulation · Discount Rate Formulation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Always `null`: costs are reported per stage.
`total_cost`	Float64	No	USD	The stage objective, `immediate_cost` + `future_cost`, in this stage’s monetary units (not discounted to the first stage).
`immediate_cost`	Float64	No	USD	The stage objective without the future-cost term, in this stage’s monetary units.
`future_cost`	Float64	No	USD	The future-cost term of the stage objective: the future-cost variable times this stage’s one-step discount factor (the single-stage factor, not the cumulative `discount_factor`). See Discount Rate Formulation.
`discount_factor`	Float64	No	—	Cumulative discount factor of this stage: multiplying a cost of this stage by it gives its present value at the first stage. See Discount Rate Formulation.
`thermal_cost`	Float64	No	USD	Thermal generation cost component.
`anticipated_thermal_cost`	Float64	No	USD	Fuel cost of the anticipated commitments decided at this stage, in this stage’s monetary units — each commitment’s delivery-stage cost discounted from its delivery stage to this stage; multiply by `discount_factor` for its present value. Zero when no anticipated units exist.
`contract_cost`	Float64	No	USD	Energy contract cost component (positive for imports, negative for exports).
`deficit_cost`	Float64	No	USD	Cost of unserved load (deficit penalty).
`excess_cost`	Float64	No	USD	Cost of excess generation (excess penalty).
`storage_violation_cost`	Float64	No	USD	Always `0`. The storage-floor and filling-target slack costs (both exist only on a hydro with a `filling` block) are included in `immediate_cost` and in no category column.
`filling_target_cost`	Float64	No	USD	Always `0`. The storage-floor and filling-target slack costs (both exist only on a hydro with a `filling` block) are included in `immediate_cost` and in no category column.
`hydro_violation_cost`	Float64	No	USD	Sum of the outflow, turbined, generation, evaporation and withdrawal violation costs.
`outflow_violation_below_cost`	Float64	No	USD	Cost of total outflow below-minimum violations.
`outflow_violation_above_cost`	Float64	No	USD	Cost of total outflow above-maximum violations.
`turbined_violation_cost`	Float64	No	USD	Cost of the turbined-flow below-minimum slack.
`generation_violation_cost`	Float64	No	USD	Cost of the generation below-minimum slack.
`evaporation_violation_cost`	Float64	No	USD	Cost of evaporation violations.
`withdrawal_violation_cost`	Float64	No	USD	Cost of water withdrawal violations.
`inflow_penalty_cost`	Float64	No	USD	Cost of inflow non-negativity slack (numerical penalty).
`generic_violation_cost`	Float64	No	USD	Cost of generic constraint violations.
`spillage_cost`	Float64	No	USD	Cost of spillage and diversion flow.
`turbined_cost`	Float64	No	USD	Turbined-flow cost: each hydro’s per-unit turbined-flow cost rate (the `turbined_cost` input) times its turbined flow and the block hours, summed over hydros and blocks.
`curtailment_cost`	Float64	No	USD	Cost of non-controllable source curtailment.
`exchange_cost`	Float64	No	USD	Transmission exchange cost component.
`pumping_cost`	Float64	No	USD	Always `0`: pumping carries no objective cost; its power enters the bus balance.

`simulation/hydros/`

Hydro plant dispatch results. One row per (stage, block, hydro) triplet. 41 columns.

Methodology: LP Formulation · Hydro Production Function Models

The nine energy-conversion columns (equivalent_productivity_mw_per_m3s … stored_energy_final_mwh, plus the four tail columns below) report the two evaluators defined in Hydro Production Function Models §5: the reference-point pair and the useful-range mean (integrated_*) pair.

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`hydro_id`	Int32	No	—	Hydro plant ID.
`turbined_m3s`	Float64	No	m³/s	Turbined flow in cubic metres per second (m³/s).
`spillage_m3s`	Float64	No	m³/s	Spilled flow in m³/s.
`outflow_m3s`	Float64	No	m³/s	Total outflow (turbined + spilled) in m³/s.
`evaporation_m3s`	Float64	Yes	m³/s	Net evaporation flow in m³/s; signed. Positive values are net evaporative loss; negative values are net rainfall on the lake surface. On a parallel stage the stage-level value, repeated on every block row; on a chronological stage the block’s own value. `0.0` for a plant that models no evaporation. Never null in written output.
`diverted_inflow_m3s`	Float64	Yes	m³/s	Flow diverted into this reservoir by other plants’ diversions, in m³/s; `0.0` when none. Never null in written output.
`diverted_outflow_m3s`	Float64	Yes	m³/s	Flow this plant diverts, in m³/s; `0.0` when it has no diversion. Never null in written output.
`incremental_inflow_m3s`	Float64	No	m³/s	Natural incremental inflow to this reservoir in m³/s (excluding upstream contributions).
`inflow_m3s`	Float64	No	m³/s	Equal to `incremental_inflow_m3s`: the plant’s incremental (local) inflow in m³/s, the stage value repeated on every block row; upstream releases are not included.
`storage_initial_hm3`	Float64	No	hm³	Reservoir storage at the start of the row’s interval in cubic hectometres (hm³): the stage’s incoming storage on a parallel stage (the same on every block row), the block’s starting boundary storage on a chronological stage.
`storage_final_hm3`	Float64	No	hm³	Reservoir storage at the end of the row’s interval in hm³: the stage’s end storage on a parallel stage, the block’s ending boundary storage on a chronological stage.
`generation_mw`	Float64	No	MW	Average power generation over the block in megawatts (MW).
`generation_mwh`	Float64	No	MWh	Total energy generated over the block in megawatt-hours (MWh).
`equivalent_productivity_mw_per_m3s`	Float64	No	MW/(m³/s)	Equivalent productivity ρ_eq [MW/(m³/s)] at the reference operating point for this stage.
`accumulated_productivity_mw_per_m3s`	Float64	No	MW/(m³/s)	Accumulated cascade productivity ρ_acum [MW/(m³/s)]: sum of ρ_eq for this plant and all downstream plants.
`incremental_inflow_energy_mw`	Float64	No	MW	Power equivalent of incremental inflow: ρ_acum × incremental_inflow_m3s [MW].
`stored_energy_initial_mwh`	Float64	No	MWh	Energy content of usable storage at stage start: (storage_initial_hm3 − V_min) × integrated_accumulated_productivity_mw_per_m3s × 1e6/3600 [MWh], where `V_min` is the plant’s physical `reservoir.min_storage_hm3` (never a `hydro_bounds` storage override) and `storage_initial_hm3` is the row’s own value (the per-block boundary storage on a chronological stage).
`stored_energy_final_mwh`	Float64	No	MWh	Energy content of usable storage at stage end: (storage_final_hm3 − V_min) × integrated_accumulated_productivity_mw_per_m3s × 1e6/3600 [MWh], where `V_min` is the plant’s physical `reservoir.min_storage_hm3` (never a `hydro_bounds` storage override) and `storage_final_hm3` is the row’s own value (the per-block boundary storage on a chronological stage).
`spillage_cost`	Float64	No	USD	Monetary cost attributed to spillage.
`water_value_per_hm3`	Float64	No	USD/hm³	Dual of the plant’s own water-balance row, in monetary units per hm³: each block’s own row on a chronological stage; the stage row, repeated on every block row, on a parallel stage. Negative values mean one more hm³ lowers the cost; see Convergence & Diagnostics for the sign.
`storage_binding_code`	Int8	No	—	Storage-binding code (see `codes.json` `storage_binding` mapping); always `0` in written output.
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `1` in written output.
`turbined_slack_m3s`	Float64	No	m³/s	Turbined flow slack variable (non-negativity enforcement). Zero under normal operation.
`outflow_slack_below_m3s`	Float64	No	m³/s	Outflow lower-bound slack in m³/s.
`outflow_slack_above_m3s`	Float64	No	m³/s	Outflow upper-bound slack in m³/s.
`generation_slack_mw`	Float64	No	MW	Generation bound slack in MW.
`storage_violation_below_hm3`	Float64	No	hm³	Storage below the dead volume of a filling hydro at its Operating stages, in hm³ (the soft floor of Penalty System); `0` on every other plant and stage: a non-filling plant in service keeps its dead volume as a hard bound, a filling plant’s floor while it fills is its filling target (whose miss is `filling_target_violation_hm3`), and any other plant outside its commissioning window has no storage floor. A stage-level value, repeated on every block row.
`filling_target_violation_hm3`	Float64	No	hm³	Filling target miss in hm³. Zero when the target is met. A stage-level value, repeated on every block row.
`evaporation_violation_pos_m3s`	Float64	No	m³/s	Slack of the evaporation row above its linearised target (evaporation exceeds the target), in m³/s; on a parallel stage the stage-level value, repeated on every block row; on a chronological stage the block’s own. Zero under normal operation.
`evaporation_violation_neg_m3s`	Float64	No	m³/s	Slack of the evaporation row below its linearised target (evaporation falls short of the target), in m³/s; on a parallel stage the stage-level value, repeated on every block row; on a chronological stage the block’s own. Zero under normal operation.
`inflow_nonnegativity_slack_m3s`	Float64	No	m³/s	Inflow non-negativity slack in m³/s. Zero under normal operation. A stage-level value, repeated on every block row.
`water_withdrawal_violation_pos_m3s`	Float64	No	m³/s	Water withdrawal over-target violation in m³/s. Zero when withdrawal is at or below target. A stage-level value, repeated on every block row.
`water_withdrawal_violation_neg_m3s`	Float64	No	m³/s	Water withdrawal under-target violation in m³/s. Zero when withdrawal is at or above target. A stage-level value, repeated on every block row.
`integrated_equivalent_productivity_mw_per_m3s`	Float64	No	MW/(m³/s)	Useful-range mean equivalent productivity [MW/(m³/s)]: ρ_eq with the forebay level averaged over the plant’s physical storage range (`reservoir` in `system/hydros.json`). Equals `equivalent_productivity_mw_per_m3s` when the range is collapsed, the plant has no VHA geometry or specific productivity, or an `equivalent_productivity_mw_per_m3s` override is supplied.
`integrated_accumulated_productivity_mw_per_m3s`	Float64	No	MW/(m³/s)	`integrated_equivalent_productivity_mw_per_m3s` summed over this plant and every downstream plant [MW/(m³/s)].
`stored_energy_initial_mw`	Float64	No	MW	`stored_energy_initial_mwh` ÷ the stage’s total block hours [MW] — the same divisor on every block row of the stage, never the block’s own hours.
`stored_energy_final_mw`	Float64	No	MW	`stored_energy_final_mwh` ÷ the stage’s total block hours [MW].

The stored_energy_initial_mwh and stored_energy_final_mwh columns equal the reference-point product (with accumulated_productivity_mw_per_m3s) only when a plant’s physical storage range is collapsed or it has no resolvable VHA geometry.

`simulation/hydro_bus_generation/`

Hydro dispatch results at (hydro, bus) cell granularity — one cell per distinct bus among a plant’s unit groups (see hydros[].unit_groups[] in the Case Format reference). One row per (stage, block, hydro, bus) quadruplet. 9 columns.

Methodology: LP Formulation

simulation/hydros/ reports each plant’s total: a plant split across several unit groups sharing one bus has no per-group quantity to report, since every split of that shared cell’s flow across its same-bus groups is an equally optimal solution with no dual to distinguish them. This partition reports at the bus-cell granularity the LP itself solves.

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`hydro_id`	Int32	No	—	Hydro plant ID.
`bus_id`	Int32	No	—	Bus ID for this cell. Always non-null.
`turbined_m3s`	Float64	No	m³/s	Turbined flow for this cell in m³/s.
`generation_mw`	Float64	No	MW	Average power generation for this cell over the block in MW.
`generation_mwh`	Float64	No	MWh	Total energy generated by this cell over the block in MWh (`generation_mw × block_duration_hours`).

A plant’s cell rows sum bit-exactly to its simulation/hydros/ row for turbined_m3s, on every production model. For generation_mw the cell rows sum bit-exactly only for FPHA and single-cell plants, and for generation_mwh only for single-cell plants, since each row multiplies its own generation_mw by the block duration. On a constant-productivity ("constant_productivity"/"linearized_head") model a multi-cell plant’s total is (Σ_c q_c)·ρ — its cells’ turbined flow summed, then multiplied by the productivity once — while each cell row is q_c·ρ computed independently. The sum of a plant’s cell rows can therefore differ from its plant total by a rounding-scale amount (see Determinism & Provenance — Out of Scope).

`simulation/thermals/`

Thermal unit dispatch results. One row per (stage, block, thermal) triplet. 12 columns.

Methodology: Equipment-Specific Formulations

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`thermal_id`	Int32	No	—	Thermal unit ID.
`generation_mw`	Float64	No	MW	Average power generation over the block in MW.
`generation_mwh`	Float64	No	MWh	Total energy generated over the block in MWh.
`generation_cost`	Float64	No	USD	Monetary generation cost for this block.
`is_anticipated`	Boolean	No	—	`true` if this unit is configured for anticipated dispatch.
`anticipated_committed_mw`	Float64	Yes	MW	Committed capacity under anticipated dispatch in MW. `null` for non-anticipated units.
`anticipated_decision_mw`	Float64	Yes	MW	Dispatch decision under anticipated dispatch in MW. `null` for non-anticipated units.
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `1` in written output.

`simulation/exchanges/`

Transmission line flow results. One row per (stage, block, line) triplet. 13 columns.

Methodology: Equipment-Specific Formulations

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`line_id`	Int32	No	—	Transmission line ID.
`direct_flow_mw`	Float64	No	MW	Flow in the forward (direct) direction in MW.
`reverse_flow_mw`	Float64	No	MW	Flow in the reverse direction in MW.
`net_flow_mw`	Float64	No	MW	Net flow (direct minus reverse) in MW.
`net_flow_mwh`	Float64	No	MWh	Net energy flow over the block in MWh.
`losses_mw`	Float64	No	MW	Transmission losses in MW.
`losses_mwh`	Float64	No	MWh	Transmission losses in MWh over the block.
`exchange_cost`	Float64	No	USD	Monetary cost attributed to this line’s exchange.
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `2` in written output.

`simulation/buses/`

Bus load balance results. One row per (stage, block, bus) triplet. 12 columns.

Methodology: System Element Modeling Overview

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`bus_id`	Int32	No	—	Bus ID.
`load_mw`	Float64	No	MW	Total load demand at this bus in MW.
`load_mwh`	Float64	No	MWh	Total load energy demand over the block in MWh.
`deficit_mw`	Float64	No	MW	Unserved load (deficit) at this bus in MW. Zero under feasible dispatch.
`deficit_mwh`	Float64	No	MWh	Unserved load energy over the block in MWh.
`excess_mw`	Float64	No	MW	Excess generation at this bus in MW. Zero under feasible dispatch.
`excess_mwh`	Float64	No	MWh	Excess generation energy over the block in MWh.
`spot_price`	Float64	No	USD/MWh	Locational marginal price (shadow price of the power balance constraint) in USD per MWh.

`simulation/pumping_stations/`

Pumping station results. One row per (stage, block, pumping station) triplet. 11 columns.

Methodology: Equipment-Specific Formulations

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`pumping_station_id`	Int32	No	—	Pumping station ID.
`pumped_flow_m3s`	Float64	No	m³/s	Pumped flow rate in m³/s.
`pumped_volume_hm3`	Float64	No	hm³	Total pumped volume over the block in hm³.
`power_consumption_mw`	Float64	No	MW	Power consumed by the pumping station in MW.
`energy_consumption_mwh`	Float64	No	MWh	Energy consumed over the block in MWh.
`pumping_cost`	Float64	No	USD	Always `0`: pumping carries no objective cost; its power enters the bus balance.
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `1` in written output.

`simulation/contracts/`

Energy contract results. One row per (stage, block, contract) triplet. 10 columns.

Methodology: Equipment-Specific Formulations

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`contract_id`	Int32	No	—	Contract ID.
`power_mw`	Float64	No	MW	Contracted power in MW, non-negative for both import and export contracts. Direction is carried by the contract type and the price sign, not by the sign of this value.
`energy_mwh`	Float64	No	MWh	Contracted energy over the block in MWh.
`price_per_mwh`	Float64	No	USD/MWh	Contract price in USD per MWh.
`total_cost`	Float64	No	USD	Total contract cost for this block: positive for imports (cost), negative for exports (revenue).
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `1` for contracts (a dormant stage emits a zero-`power_mw` row, not a distinct code).

`simulation/non_controllables/`

Non-controllable source results (wind, solar, run-of-river hydro without storage, etc.). One row per (stage, block, non-controllable) triplet. 12 columns.

Methodology: Equipment-Specific Formulations

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. Nullable in the schema; always set in written output.
`non_controllable_id`	Int32	No	—	Non-controllable source ID.
`generation_mw`	Float64	No	MW	Actual generation dispatched in MW.
`generation_mwh`	Float64	No	MWh	Actual energy generated over the block in MWh.
`available_mw`	Float64	No	MW	Maximum available generation in MW (before curtailment).
`curtailment_mw`	Float64	No	MW	Generation curtailed in MW. Zero when all available generation is dispatched.
`curtailment_mwh`	Float64	No	MWh	Curtailed energy over the block in MWh.
`curtailment_cost`	Float64	No	USD	Monetary cost attributed to curtailment.
`operative_state_code`	Int8	No	—	Operative state code (see `codes.json` `operative_state` mapping); always `1` in written output.

`simulation/inflow_lags/`

Autoregressive inflow lag state variables. One row per (stage, hydro, lag) triplet. No block dimension — inflow lags are stage-level state variables. 6 columns. All columns are non-nullable.

Methodology: PAR(p) Inflow Model · State Augmentation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`hydro_id`	Int32	No	—	Hydro plant ID.
`lag_index`	Int32	No	—	Autoregressive lag index, 0-based: `0` is the inflow of the most recently completed lag period (Multi-Resolution Studies §2); without `season_definitions` no stage completes a lag period and `inflow_m3s` is `0` on every row.
`inflow_m3s`	Float64	No	m³/s	Inflow value for this lag in m³/s.

`simulation/in_transit/`

Water travel-time in-transit volumes. One row per (stage, downstream hydro, maturity lag) triplet. No block dimension — in-transit buckets are stage-level state variables. 7 columns. All columns are non-nullable. The directory is present only when the system declares a travel-time arc (travel_time_hours present, strictly positive, with a downstream_id).

Methodology: State Augmentation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`hydro_id`	Int32	No	—	Downstream (receiving) hydro plant ID the in-transit water is destined for.
`lag`	Int32	No	—	Maturity lag (1-based) of the bucket at the end of this stage: lag 1 arrives at the receiving plant at the next stage, lag `k` arrives `k` stages later.
`in_transit_volume_hm3`	Float64	No	hm³	Volume in this bucket at the end of the stage, in hm³. Water that would mature after the last stage is dropped when no terminal boundary is loaded, so such a bucket reads `0` (see Horizon limitation).
`delayed_arrival_hm3`	Float64	No	hm³	Volume of the incoming lag-1 bucket, which enters the receiving plant’s water balance at this stage or, while that plant is outside its commissioning window, the balance of the nearest downstream plant that is not PreFilling (with none, it leaves the modeled system), in hm³; written on the `lag` = 1 row and `0` on deeper lags.

`simulation/transit_seed/`

Rolling release-window seed recording each in-transit bucket’s own upstream release history, for seeding a continuing (resumed) run’s transit buckets across the boundary. Scenario-level — unlike every other simulation partition, a window’s own [start_date, end_date) span anchors the row, not a stage/node index, so this schema carries scenario_id alone (no stage_id/node_id). 5 columns. All columns are non-nullable. Hive-partitioned by scenario_id (transit_seed/scenario_id=NNNN/data.parquet); written only when the system declares a travel-time arc.

Methodology: State Augmentation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`hydro_id`	Int32	No	—	Upstream entity identifier whose release the window covers.
`start_date`	Date32	No	—	Start of the release window (inclusive).
`end_date`	Date32	No	—	End of the release window (exclusive).
`value_m3s`	Float64	No	m³/s	Mean release rate over the window, in m³/s.

`simulation/anticipated_lanes/`

Post-horizon commitments the study decides, keyed (thermal_id, delivery_date) — distinct from the per-plant anticipated_committed_mw/anticipated_decision_mw columns on simulation/thermals/. One row per scenario and decision whose delivery stage is a post-study stage, at its decision stage; a scenario with no such decision writes no partition. 7 columns. All columns are non-nullable. Hive-partitioned by scenario_id (anticipated_lanes/scenario_id=NNNN/data.parquet); written only when the study declares post_study_stages.

Methodology: State Augmentation

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id of the decision stage, where the commitment is deposited into its ring slot.
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`thermal_id`	Int32	No	—	Anticipated thermal unit ID.
`delivery_date`	Int32	No	—	Calendar date the commitment delivers, encoded `YYYYMMDD` (`year10000 + month100 + day`).
`deposited_decision_mw`	Float64	No	MW	Commitment decided at `stage_id` for delivery at `delivery_date`, in MW.
`carried_committed_mw`	Float64	No	MW	Value of the ring slot holding the commitment at `stage_id`, in MW; the deposit row makes it equal to `deposited_decision_mw`.

`anticipated/fixed_deliveries.parquet`

Run-level echo of every commitment decided before the study and delivered past the study horizon (for the equivalent term in other planning tools, see the Glossary), declared as past_anticipated_commitments windows extending past the horizon end. Unlike simulation/anticipated_lanes/, this file is not under simulation/, is unpartitioned, and carries no scenario, stage, or node axis: each row is a resolved input, scenario- and stage-independent. One row per fixed post-horizon window. 4 columns. All columns are non-nullable. Written by both the CLI and the Python bindings; written only when at least one such window exists — an empty set writes no file and no anticipated/ directory. There is no cost column: the fuel of a commitment decided before the study is sunk and enters no objective; when a boundary is loaded, the commitment’s state contribution is folded into the boundary cuts’ intercepts.

Methodology: State Augmentation

Name	Type	Nullable	Units	Description
`thermal_id`	Int32	No	—	Anticipated thermal unit ID.
`start_date`	Date32	No	—	First delivery date of the fixed window.
`end_date`	Date32	No	—	End of the fixed window (exclusive): the window’s `end_date` as declared in `past_anticipated_commitments`.
`value_mw`	Float64	No	MW	Committed delivery held constant over the window, in MW.

`simulation/violations/generic/`

Generic user-defined constraint slacks. One row per active generic-constraint row at each stage — one per (stage, block, constraint) triplet, or one stage-level row (block_id null) for a block-independent constraint — whether or not it is violated; slack_value and slack_cost are 0 where it is not and on every row of a constraint with slack disabled. 7 columns.

Methodology: Penalty System

Name	Type	Nullable	Units	Description
`scenario_id`	Int32	No	—	Scenario id (0-based). Duplicates the Hive partition directory (`scenario_id=NNNN/`) as an explicit in-file column.
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`node_id`	Int32	No	—	Visited policy-graph node id; on a stage chain, the 0-based stage position (not the declared stage id).
`block_id`	Int32	Yes	—	Load block index. `null` for stage-level constraints.
`constraint_id`	Int32	No	—	Constraint ID as defined in the case input files.
`slack_value`	Float64	No	—	Non-negative for a one-sided constraint. For a two-sided (range) constraint (both bounds finite), the signed net `s_plus - s_minus` — may be negative. Zero means no violation.
`slack_cost`	Float64	No	USD	Monetary cost attributed to this violation.

`simulation/solver/iterations.parquet`

Identical schema to training/solver/iterations.parquet. One row per scenario_id; phase is always "simulation". iteration, stage_id, opening_index, rank, and worker_id are NULL (a simulation row carries no per-iteration / per-stage / per-opening / per-rank / per-worker attribution); scenario_id is set.

Methodology: LP Warm-Start

`simulation/solver/retry_histogram.parquet`

Same columns and types as training/solver/retry_histogram.parquet. One row per scenario and retry level with a positive count; iteration holds the scenario id (0-based), phase is always "simulation" and stage_id is null. novomodelo run (CLI) writes the file with no rows and records simulation retries only in lp_retries and retry_attempts of simulation/solver/iterations.parquet. The Python function novomodelo.run.run() writes the rows.

`generic_constraints/resolved_echo.parquet`

The generic_constraints/ directory is written when the study declares any generic user-defined constraints. It sits at the top level of <output_dir>/, a peer of training/ and simulation/, not nested under either. The directory is omitted entirely when the study has no generic constraints.

Methodology: LP Formulation

The fully resolved echo of every generic constraint as the LP actually built it — one row per (constraint, stage, block, term). Written once, not per-scenario (the resolution is deterministic given the case, independent of simulation outcome). 13 columns.

bound_lower/bound_upper are the resolved interval endpoints (lower before upper), each null where unbounded on that side; derived_shape labels the shape those endpoints imply: floor, cap, band or equality. The per-term columns (term_index, variable_kind, variable, coefficient) are null on a term-less constraint’s placeholder row, and slack_penalty is null when slack is disabled for the constraint.

Name	Type	Nullable	Units	Description
`stage_id`	Int32	No	—	Declared stage id (from `stages.json`).
`block_id`	Int32	Yes	—	Load block index within the stage. `null` for stage-level (non-block) constraints.
`constraint_id`	Int32	No	—	Generic constraint ID as defined in the case input files.
`constraint_name`	Utf8	No	—	Human-readable constraint name from the case input files.
`term_index`	Int32	Yes	—	Index of this term within the constraint’s linear expression. `null` on a term-less placeholder row.
`variable_kind`	Utf8	Yes	—	Kind of decision variable this term references. `null` on a term-less placeholder row.
`variable`	Utf8	Yes	—	Identity of the referenced variable within its kind. `null` on a term-less placeholder row.
`coefficient`	Float64	Yes	—	Linear coefficient applied to this term. `null` on a term-less placeholder row.
`bound_lower`	Float64	Yes	—	Resolved lower bound of the constraint’s interval. `null` when unbounded below.
`bound_upper`	Float64	Yes	—	Resolved upper bound of the constraint’s interval. `null` when unbounded above.
`derived_shape`	Utf8	No	—	Shape implied by the resolved bounds: `floor`, `cap`, `band` or `equality`.
`slack_enabled`	Boolean	No	—	Whether a slack variable is enabled for this constraint.
`slack_penalty`	Float64	Yes	—	Penalty cost per unit of slack. `null` when slack is disabled.