Flow Preprocessing

Flow Preprocessing#

This page documents the classes and functions in the flow_preprocessing module of the HydroGenerate package.

class FlowPreProcessingParameters(minimum_turbineflow, minimum_turbineflow_percent, annual_maintenance_flag, major_maintenance_flag)#

Bases: object

Configuration container for flow preprocessing.

This class holds user-configurable parameters that control minimum turbine flow enforcement and maintenance scheduling. It does not perform any computations by itself; the values are consumed by FlowPreProcessing.

Parameters:

  • minimum_turbineflow (float or None) – Absolute minimum turbine flow (m³/s). If provided, this overrides the percentage-based minimum flow. If None, the minimum is computed from minimum_turbineflow_percent (or a default of 10%).

  • minimum_turbineflow_percent (float or int or None) – Minimum turbine flow as a percentage of design_flow (1–100). Used only when minimum_turbineflow is None.

  • annual_maintenance_flag (bool) – If True, apply annual maintenance by setting turbine flow to zero for the lowest-flow calendar week each year (based on mean weekly flow).

  • major_maintenance_flag (bool) – If True, apply major maintenance by setting turbine flow to zero for the lowest-flow two-week period every five years.

Notes

These parameters are typically attached to a higher-level flow object (e.g., a project-specific container) and then read by methods in FlowPreProcessing.

class FlowPreProcessing#

Bases: object

Flow preprocessing routines operating on a provided flow object.

Methods in this class update values on flow_obj in-place (most importantly flow_obj.turbine_flow).

Notes

The caller is responsible for ensuring that:

  • flow_obj.turbine_flow is a 1D numpy array aligned with flow_obj.datetime_index.

  • flow_obj.datetime_index is a pandas.DatetimeIndex whose name (or level name) is "dateTime" so that pd.Grouper(level="dateTime", ...) works as written in the implementation.

max_turbineflow_checker(flow_obj)#

Cap turbine flow at the design flow.

Parameters:

flow_obj (object) –

A flow object with attributes:

  • turbine_flow (numpy.ndarray)

  • design_flow (float)

  • minimum_turbineflow (float or None)

  • minimum_turbineflow_percent (float or None)

Notes

Updates flow_obj.turbine_flow (numpy.ndarray), where each element is min(flow, design_flow).

min_turbineflow_checker(flow_obj)#

Enforce a minimum turbine flow threshold.

Any values in flow_obj.turbine_flow that fall below the computed minimum are replaced with zero.

Parameters:

flow_obj (object)

Notes

This method modifies flow_obj.turbine_flow in place.

If flow_obj.minimum_turbineflow is None, the minimum flow is computed as:

min_flow = (minimum_turbineflow_percent / 100) * design_flow

where minimum_turbineflow_percent defaults to 10 if not provided.

annual_maintenance_implementer(flow_obj)#

Apply annual maintenance by zeroing the lowest-flow week each year.

The method identifies the calendar week with the lowest mean weekly flow (aggregated across all years) and sets turbine flow to zero for that week in every year of the dataset.

Parameters:

flow_obj (object)

Notes

This method modifies flow_obj.turbine_flow in place.

Weekly means are computed using freq='W' and grouped by calendar week number (%W). The maintenance window spans 7 days starting on the Monday of the selected week.

major_maintenance_implementer(flow_obj)#

Apply major maintenance by zeroing a low-flow two-week period every 5 years.

The method identifies the calendar week associated with the lowest mean two-week flow and applies a 14-day zero-flow window starting from that week in each major-maintenance year.

Parameters:

flow_obj (object)

Notes

This method modifies flow_obj.turbine_flow in place.

Major maintenance years occur every five years starting after the first year of the dataset. Two-week means are computed using freq='2W'.

Contents