Page 261 - Intelligent Digital Oil And Gas Fields

Integrated Asset Management and Optimization Workflows 211

Table 6.1 Selected (Meta)heuristic Optimization Methods With Main
Applications—cont’d
Optimization
Metaheuristic Technique Application Reference
Markov chain Monte Carlo AHM Schulze-Riegert et al.
(McMC) (2016), Maucec et al.
(2007, 2011, 2013a,b),
and Olalotiti-Lawal and
Datta-Gupta (2015) c
Reservoir Li and Reynolds (2017)
description and
forecasting
Prediction Fillacier et al. (2014) and
d
uncertainty Goodwin et al. (2017)
quantification
Differential evolution (DE) AHM Hajizadeh et al. (2010) and
Olalotiti-Lawal and
c
Datta-Gupta (2015)
Tabu search (TS) and scatter Multiple-field Cullick et al. (2003)
search (SS) scheduling
optimization
Project portfolio April et al. (2003)
optimization
Well-placement Cullick et al. (2006)
optimization
AHM Yang et al. (2007)
e
Artificial lift Vasquez et al. (2001)
optimization
a
HGA: hybrid technique of GA and ANNs.
b
PSO-MADS: hybrid technique of PSO and mesh adaptive direct search (MADS).
c
DEMC: hybrid technique of differential evolution (DE) and McMC.
d
HMcMC: Hamiltonian McMC.
e
GATS: hybrid technique of GA and Tabu search (TS).
expected decision error (EDE), through selecting the wrong projects”.
McVay and Dossary (2014) present a new framework for assessing the
impact of overconfidence and directional bias on portfolio or asset perfor-
mance. They further report that for moderate amounts of overconfidence
and optimism, the ED amounted to 30%–35% of NPV for analyzed portfo-
lios and optimization cases, which can profoundly affect the asset perfor-
mance. In even broader context, Allen (2017) describes handling risk and
uncertainty in portfolio and/or asset production forecasting. He builds on
portfolio optimization under uncertainty and introduces sequencing of
uncertainty and aggregation of risk as fundamental components in an asset’s
production vulnerability and associated risk management.