Merge pull request #13 from IdahoLabResearch/constraints_check

Constraints and turbine selection

.gitignore

@@ -1,7 +1,10 @@
 .DS_Store
+**/__pycache__/
+*.pyc
+__pycache__
 HydroGenerate.egg-info/
-hat/__pycache__/*
-hat/__pycache__/*
-hat/__pycache__/*
-hat/__pycache__/*
-hat/utils/__pycache__/*
+HydroGenerate/__pycache__/*
+HydroGenerate/__pycache__/*
+HydroGenerate/__pycache__/*
+HydroGenerate/__pycache__/*
+HydroGenerate/utils/__pycache__/*

HydroGenerate/flow_preprocessing.py

@@ -0,0 +1,133 @@
+'''
+Copyright 2023, Battelle Energy Alliance, LLC
+'''
+
+"""
+07-2024 
+@author: Soumyadeep Nag, Camilo J. Bastidas Pacheco, J. Gallego-Calderon
+
+This module applies the flow limit constaints and incorporates the impact of annual maintenance activities
+"""
+
+import numpy as np
+import pandas as pd
+from datetime import datetime
+
+
+# Hydraulic design parameters
+class FlowPreProcessingParameters:
+    def __init__(self, minimum_turbineflow, minimum_turbineflow_percent,       # Flow parameters
+                 annual_maintenance_flag, major_maintenance_flag): 
+        '''
+#         This class initializes Flow preprocessing parameters needed for multiple calculations 
+#         Parameter descriptions are provided below:
+#         '''         
+# Inputs
+        self.minimum_turbineflow = minimum_turbineflow        # minimum flow through the turbine (m3/s)     
+        self.minimum_turbineflow_percent = minimum_turbineflow_percent      # Set minimum flow by using a percent of design flow, deault is 10  (%, 1-100)
+        self.annual_maintenance_flag = annual_maintenance_flag        # Boolean to include annual maintenance or not
+        self.major_maintenance_flag = major_maintenance_flag      # Boolean to include major maintenance or not
+
+
+
+class FlowPreProcessing():
+
+    # Function that sets max turbine flow to the design flow
+    def max_turbineflow_checker(self, flow_obj):
+
+        flow = flow_obj.flow            # this is a numpy array
+        design_flow = flow_obj.design_flow
+        flow_obj.turbine_flow = np.where(flow > design_flow, design_flow, flow)         # turbine_flow is created here as this code is always active and runs 1st. flow is < design flow
+
+    # Function that sets min turbine flow to the design flow
+    def min_turbineflow_checker(self, flow_obj):
+        flow = flow_obj.turbine_flow # this is a numpy array        
+        design_flow = flow_obj.design_flow
+
+        # check if the user entered a min flow value
+        if flow_obj.minimum_turbineflow is not None:
+            min_flow = flow_obj.minimum_turbineflow
+        else:
+            if flow_obj.minimum_turbineflow_percent is not None:         # Check if a user entered min flow percentage if not, use 10% as default
+                min_flow_percent =  flow_obj.minimum_turbineflow_percent
+            else:
+                min_flow_percent = 10           # Default   
+
+            min_flow = (min_flow_percent / 100) * design_flow           # Compute minimum flow that passes through the turbine
+            flow_obj.minimum_turbineflow = min_flow            # update
+
+        flow_obj.turbine_flow = np.where(flow < min_flow, 0, flow)          # replace flows less than min flow with 0
+
+    def annual_maintenance_implementer(self, flow_obj):
+        # Function to set annual maintennace - i.e., make the flow 0 for a week a year
+
+        turbine_flow = flow_obj.turbine_flow            # turbine flow series with max and min (if on) limits implemented
+        datetime_index = flow_obj.datetime_index            # datetime index of the user flow
+        flow = pd.DataFrame(data= turbine_flow, index= datetime_index, columns=['flow_cms'])            # create dataframe with date and flow in m3/s
+
+        # Create series with maintennance years
+        start_date = datetime_index[0].year         # Start Year
+        end_date = datetime_index[-1].year          # End year
+        maint_years = list(range(start_date, end_date, 1))          # Years where there will be maintenance - every year
+
+        # Find minimum week - maintennance week
+        weekly = flow .groupby(pd.Grouper(level="dateTime", freq='W')).mean().iloc[1:-1]             # compute the two-weekly mean and drop first and last
+        weekly['week'] = weekly.index.strftime('%W')
+        week_maint = weekly.groupby('week').mean().idxmin()[0]
+
+        # Generate a series with the monday of every maintenance week
+        start_days = []
+        for m in maint_years:
+            start_days.append(datetime.strptime(str(m) + str(week_maint) + '1', "%Y%W%w"))          # Monday of the week where maintenance starts
+
+        # Generate a date series of maintenance from start date
+        dates_maint = []        # maintenance dates
+        for d in start_days:
+            # for day in range(7):
+            dates_maint.append(pd.date_range(d, periods= 7).date)           # generate 7 days for each year
+            #  dates_maint.append((sd + datetime.timedelta(days=day)).strftime("%Y-%m-%d"))
+
+        dates_maint = np.concatenate(dates_maint, axis= 0)      # generate single array
+        flow['date'] = flow.index.date          # add date to dataframe
+        flow.loc[flow.date.isin(dates_maint), 'flow_cms'] = 0           # replace flow in minimum weeks with 0
+
+        flow_obj.turbine_flow = flow['flow_cms'].to_numpy()         # update
+
+    # Function that schedules the major maintennace
+    def major_maintenance_implementer(self, flow_obj):
+        # Major maintenance will happend on the month with lowest flow a year - during the first 15 days of this month.
+
+        turbine_flow = flow_obj.turbine_flow            # turbine flow series with max and min (if on) limits implemented
+        datetime_index = flow_obj.datetime_index            # datetime index of the user flow
+        flow = pd.DataFrame(data= turbine_flow, index= datetime_index, columns=['flow_cms']) # create dataframe with date and flow in m3/s
+
+        # Find maintennance years
+        start_date = datetime_index[0].year         # Start Year
+        end_date = datetime_index[-1].year          # End year
+        maint_years = list(range(start_date, end_date, 5))          # Years where there will be maintenance - every 5 years
+        del maint_years[0]          # remove the first year
+
+        # Find minimum two-week period - maintennance week
+        weekly = flow .groupby(pd.Grouper(level="dateTime", freq='2W')).mean().iloc[1:-1]             # compute the two-weekly mean and drop first and last
+        weekly['week'] = weekly.index.strftime('%W')
+        week_maint = weekly.groupby('week').mean().idxmin()[0]
+
+        # Generate a series with the monday of every maintenance week
+        start_days = []
+        for m in maint_years:
+            start_days.append(datetime.strptime(str(m) + str(week_maint) + '1', "%Y%W%w"))          # Monday of the week where maintenance starts
+
+        # Generate a date series of maintenance from start date
+        dates_maint = []
+        for d in start_days:
+            dates_maint.append(pd.date_range(d, periods= 14).date)           # generate 14 days for each year
+
+        dates_maint = np.concatenate(dates_maint, axis= 0)
+
+        # Make flow 0 during those days
+        flow['date'] = flow.index.date
+        flow.loc[flow.date.isin(dates_maint), 'flow_cms'] = 0           # replace flow in minimum weeks with 0
+
+        flow_obj.turbine_flow = flow['flow_cms'].to_numpy()         # update
+
+