Optimizing vessel fleet size and mix to support maintenance operations at offshore wind farms

Magnus Stålhane; Elin E. Halvorsen-Weare; Lars Magne Nonås; Giovanni Pantuso

doi:10.1016/j.ejor.2019.01.023

Optimizing vessel fleet size and mix to support maintenance operations at offshore wind farms

Magnus Stålhane^*, Elin E. Halvorsen-Weare, Lars Magne Nonås, Giovanni Pantuso

^*Corresponding author for this work

Department of Mathematical Sciences

7 Citations (Scopus)

Abstract

This paper considers the problem of determining the optimal vessel fleet to support maintenance operations at an offshore wind farm. We propose a two-stage stochastic programming (SP) model of the problem where the first stage decisions are what vessels to charter. The second stage decisions are how to support maintenance tasks using the chartered vessels from the first stage, given uncertainty in weather conditions and the occurrence of failures. To solve the resulting SP model we perform an ad-hoc Dantzig–Wolfe decomposition where, unlike standard decomposition approaches for SP models, parts of the second stage problem remain in the master problem. The decomposed model is then solved as a matheuristic by apriori generating a subset of the possible extreme points from the Dantzig–Wolfe subproblems. A computational study in three parts is presented. First, we verify the underlying mathematical model by comparing results to leading work from the literature. Then, results from in-sample and out-of-sample stability tests are presented to verify that the matheuristic gives stable results. Finally, we exemplify how the model can help offshore wind farm operators and vessel developers improve their decision making processes.

Original language	English
Journal	European Journal of Operational Research
Volume	276
Issue number	2
Pages (from-to)	495-509
ISSN	0377-2217
DOIs	https://doi.org/10.1016/j.ejor.2019.01.023
Publication status	Published - 16 Jul 2019

Keywords

Fleet size and mix
Logistics
Maintenance planning
Offshore wind
Stochastic programming

Access to Document

10.1016/j.ejor.2019.01.023

OA-Optimizing vesselAccepted author manuscript, 538 KBLicence: CC BY-NC-ND

Cite this

@article{8b53771362b14fc9b6338568f65a4e7a,

title = "Optimizing vessel fleet size and mix to support maintenance operations at offshore wind farms",

abstract = "This paper considers the problem of determining the optimal vessel fleet to support maintenance operations at an offshore wind farm. We propose a two-stage stochastic programming (SP) model of the problem where the first stage decisions are what vessels to charter. The second stage decisions are how to support maintenance tasks using the chartered vessels from the first stage, given uncertainty in weather conditions and the occurrence of failures. To solve the resulting SP model we perform an ad-hoc Dantzig–Wolfe decomposition where, unlike standard decomposition approaches for SP models, parts of the second stage problem remain in the master problem. The decomposed model is then solved as a matheuristic by apriori generating a subset of the possible extreme points from the Dantzig–Wolfe subproblems. A computational study in three parts is presented. First, we verify the underlying mathematical model by comparing results to leading work from the literature. Then, results from in-sample and out-of-sample stability tests are presented to verify that the matheuristic gives stable results. Finally, we exemplify how the model can help offshore wind farm operators and vessel developers improve their decision making processes.",

keywords = "Fleet size and mix, Logistics, Maintenance planning, Offshore wind, Stochastic programming",

author = "Magnus St{\aa}lhane and Halvorsen-Weare, {Elin E.} and Non{\aa}s, {Lars Magne} and Giovanni Pantuso",

year = "2019",

month = jul,

day = "16",

doi = "10.1016/j.ejor.2019.01.023",

language = "English",

volume = "276",

pages = "495--509",

journal = "European Journal of Operational Research",

issn = "0377-2217",

publisher = "Elsevier",

number = "2",

}

TY - JOUR

T1 - Optimizing vessel fleet size and mix to support maintenance operations at offshore wind farms

AU - Stålhane, Magnus

AU - Halvorsen-Weare, Elin E.

AU - Nonås, Lars Magne

AU - Pantuso, Giovanni

PY - 2019/7/16

Y1 - 2019/7/16

N2 - This paper considers the problem of determining the optimal vessel fleet to support maintenance operations at an offshore wind farm. We propose a two-stage stochastic programming (SP) model of the problem where the first stage decisions are what vessels to charter. The second stage decisions are how to support maintenance tasks using the chartered vessels from the first stage, given uncertainty in weather conditions and the occurrence of failures. To solve the resulting SP model we perform an ad-hoc Dantzig–Wolfe decomposition where, unlike standard decomposition approaches for SP models, parts of the second stage problem remain in the master problem. The decomposed model is then solved as a matheuristic by apriori generating a subset of the possible extreme points from the Dantzig–Wolfe subproblems. A computational study in three parts is presented. First, we verify the underlying mathematical model by comparing results to leading work from the literature. Then, results from in-sample and out-of-sample stability tests are presented to verify that the matheuristic gives stable results. Finally, we exemplify how the model can help offshore wind farm operators and vessel developers improve their decision making processes.

AB - This paper considers the problem of determining the optimal vessel fleet to support maintenance operations at an offshore wind farm. We propose a two-stage stochastic programming (SP) model of the problem where the first stage decisions are what vessels to charter. The second stage decisions are how to support maintenance tasks using the chartered vessels from the first stage, given uncertainty in weather conditions and the occurrence of failures. To solve the resulting SP model we perform an ad-hoc Dantzig–Wolfe decomposition where, unlike standard decomposition approaches for SP models, parts of the second stage problem remain in the master problem. The decomposed model is then solved as a matheuristic by apriori generating a subset of the possible extreme points from the Dantzig–Wolfe subproblems. A computational study in three parts is presented. First, we verify the underlying mathematical model by comparing results to leading work from the literature. Then, results from in-sample and out-of-sample stability tests are presented to verify that the matheuristic gives stable results. Finally, we exemplify how the model can help offshore wind farm operators and vessel developers improve their decision making processes.

KW - Fleet size and mix

KW - Logistics

KW - Maintenance planning

KW - Offshore wind

KW - Stochastic programming

UR - http://www.scopus.com/inward/record.url?scp=85060874312&partnerID=8YFLogxK

U2 - 10.1016/j.ejor.2019.01.023

DO - 10.1016/j.ejor.2019.01.023

M3 - Journal article

AN - SCOPUS:85060874312

SN - 0377-2217

VL - 276

SP - 495

EP - 509

JO - European Journal of Operational Research

JF - European Journal of Operational Research

IS - 2

ER -

Optimizing vessel fleet size and mix to support maintenance operations at offshore wind farms

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this