Stochastic Generation Capacity Expansion Planning with Approximate Dynamic Programming

Bryan Palmintier; Jesse Bukenberger

doi:10.1109/TDC.2018.8440368

Stochastic Generation Capacity Expansion Planning with Approximate Dynamic Programming

Bryan Palmintier, Jesse Bukenberger

Power Systems Engineering

Research output: Contribution to conference › Paper › peer-review

5 Scopus Citations

Abstract

Failing to consider the many long-term uncertainties that affect the performance of a power generation portfolio can result in suboptimal generation expansion plans. Further, traditional deterministic approaches can omit flexible plans that are able to adapt to future events. In this paper, we explore the use of approximate dynamic programming (ADP) to create forward-looking generation expansion plans. A case study is included with three sequential decision periods; three generation technologies; and four uncertainties: demand growth, natural gas prices, renewable portfolio standards, and the adoption of customer-driven solar generation. The flexible plans found through ADP show a 3 % reduction in total expected cost when compared to myopic planning heuristics while circumventing the computational burdens that accompany high-dimensional dynamic programs.

Original language	American English
Number of pages	5
DOIs	https://doi.org/10.1109/TDC.2018.8440368 https://doi.org/10.1109/TDC.2018.8440368
State	Published - 17 Aug 2018
Event	2018 IEEE/PES Transmission and Distribution Conference and Exposition, T and D 2018 - Denver, United States Duration: 16 Apr 2018 → 19 Apr 2018

Conference

Conference	2018 IEEE/PES Transmission and Distribution Conference and Exposition, T and D 2018
Country/Territory	United States
City	Denver
Period	16/04/18 → 19/04/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

NREL Publication Number

NREL/CP-5D00-69025

Keywords

Approximate Dynamic Programming
Distributed Energy Resources
Dynamic Programming
Power Generation Planning
Sampled Backward Induction
Uncertainty

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

Cite this

@conference{0ba8465037644ddfa5937e4aa9cb3f1f,

title = "Stochastic Generation Capacity Expansion Planning with Approximate Dynamic Programming",

abstract = "Failing to consider the many long-term uncertainties that affect the performance of a power generation portfolio can result in suboptimal generation expansion plans. Further, traditional deterministic approaches can omit flexible plans that are able to adapt to future events. In this paper, we explore the use of approximate dynamic programming (ADP) to create forward-looking generation expansion plans. A case study is included with three sequential decision periods; three generation technologies; and four uncertainties: demand growth, natural gas prices, renewable portfolio standards, and the adoption of customer-driven solar generation. The flexible plans found through ADP show a 3 % reduction in total expected cost when compared to myopic planning heuristics while circumventing the computational burdens that accompany high-dimensional dynamic programs.",

keywords = "Approximate Dynamic Programming, Distributed Energy Resources, Dynamic Programming, Power Generation Planning, Sampled Backward Induction, Uncertainty",

author = "Bryan Palmintier and Jesse Bukenberger",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 2018 IEEE/PES Transmission and Distribution Conference and Exposition, T and D 2018 ; Conference date: 16-04-2018 Through 19-04-2018",

year = "2018",

month = aug,

day = "17",

doi = "10.1109/TDC.2018.8440368",

language = "American English",

}

TY - CONF

T1 - Stochastic Generation Capacity Expansion Planning with Approximate Dynamic Programming

AU - Palmintier, Bryan

AU - Bukenberger, Jesse

PY - 2018/8/17

Y1 - 2018/8/17

N2 - Failing to consider the many long-term uncertainties that affect the performance of a power generation portfolio can result in suboptimal generation expansion plans. Further, traditional deterministic approaches can omit flexible plans that are able to adapt to future events. In this paper, we explore the use of approximate dynamic programming (ADP) to create forward-looking generation expansion plans. A case study is included with three sequential decision periods; three generation technologies; and four uncertainties: demand growth, natural gas prices, renewable portfolio standards, and the adoption of customer-driven solar generation. The flexible plans found through ADP show a 3 % reduction in total expected cost when compared to myopic planning heuristics while circumventing the computational burdens that accompany high-dimensional dynamic programs.

AB - Failing to consider the many long-term uncertainties that affect the performance of a power generation portfolio can result in suboptimal generation expansion plans. Further, traditional deterministic approaches can omit flexible plans that are able to adapt to future events. In this paper, we explore the use of approximate dynamic programming (ADP) to create forward-looking generation expansion plans. A case study is included with three sequential decision periods; three generation technologies; and four uncertainties: demand growth, natural gas prices, renewable portfolio standards, and the adoption of customer-driven solar generation. The flexible plans found through ADP show a 3 % reduction in total expected cost when compared to myopic planning heuristics while circumventing the computational burdens that accompany high-dimensional dynamic programs.

KW - Approximate Dynamic Programming

KW - Distributed Energy Resources

KW - Dynamic Programming

KW - Power Generation Planning

KW - Sampled Backward Induction

KW - Uncertainty

UR - http://www.scopus.com/inward/record.url?scp=85053178106&partnerID=8YFLogxK

U2 - 10.1109/TDC.2018.8440368

DO - 10.1109/TDC.2018.8440368

M3 - Paper

AN - SCOPUS:85053178106

T2 - 2018 IEEE/PES Transmission and Distribution Conference and Exposition, T and D 2018

Y2 - 16 April 2018 through 19 April 2018

ER -

Stochastic Generation Capacity Expansion Planning with Approximate Dynamic Programming

Abstract

Conference

Bibliographical note

NREL Publication Number

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this