Role of hydrologic information in stochastic dynamic programming: a case study of the Kemano hydropower system in British Columbia

This paper presents a study describing the effect of various hydrological variables in stochastic dynamic programming (SDP) for solving the optimization problem of managing a hydropower system. We will show how choosing the best hydrological variables can strongly affect management policies. This is...

Full description

Saved in:
Bibliographic Details
Published inCanadian journal of civil engineering Vol. 41; no. 9; pp. 839 - 844
Main Authors Desreumaux, Quentin, Côté, Pascal, Leconte, Robert
Format Journal Article
LanguageEnglish
Published Ottawa NRC Research Press 01.09.2014
Canadian Science Publishing NRC Research Press
Subjects
Analysis
British Columbia
Case studies
Dynamic programming
Dynamical systems
Dynamics
gestion des réservoirs
Hydroelectric power
hydrologic information
Hydrology
hydropower
hydroélectricité
information hydrologique
Management
Mathematical models
Optimization
Policies
programmation stochastique dynamique
Regression analysis
reservoir management
Reservoirs
Snow
Snow cover
Snow-water equivalent
stochastic dynamic programming
Stochastic models
Stochastic programming
Stochasticity
Stream discharge
Stream flow
Streamflow
Water-power
British Columbia
British Columbia Canada
Canada
Canada, British Columbia
Online AccessGet full text
ISSN0315-1468
1208-6029
DOI10.1139/cjce-2013-0370

Cover

More Information
Summary:This paper presents a study describing the effect of various hydrological variables in stochastic dynamic programming (SDP) for solving the optimization problem of managing a hydropower system. We will show how choosing the best hydrological variables can strongly affect management policies. This is especially true for the system studied here, namely the Kemano hydroelectric system located in British Columbia, Canada, which is subject to large streamflow volumes due to significant snow cover during winter. Real-time snow water equivalent (SWE) data can be used directly as a variable in SDP management policies. Results indicate that for the system in this study, the maximum SWE (i.e., highest level of SWE observed from the start of winter to the current decision period) is the best among the methods investigated for effective, safe management, compared with Markov or order p autoregressive models when forecasts are not available.
Bibliography:ObjectType-Case Study-2
SourceType-Scholarly Journals-1
content type line 14
ObjectType-Feature-3
ObjectType-Report-1
ObjectType-Article-1
ObjectType-Feature-2
content type line 23
ISSN:0315-1468
1208-6029
DOI:10.1139/cjce-2013-0370