Page 273 - Intelligent Digital Oil And Gas Fields

Integrated Asset Management and Optimization Workflows 223

procedure several times on the same multiple observed data. Recently, the
ES-MDA algorithms have been successfully deployed to integrated uncer-
tainty analysis to render more robust field development decisions (Hegstad
and Sætrom, 2014), reservoir modeling with integration of drill-stem (DST)
data (Sætrom et al., 2016), AHM and rigorous uncertainty quantification in
high-resolution model of a naturally fractured reservoir (Maucec et al.,
2016), as well as for the AHM with uncertainty of a large-scale, dual
porosity-dual permeability (DPDP) integrated reservoir model (IRM)
(Maucec et al., 2017) powered by a massive parallel processing simulation
platform.

6.3.4 Closed-Loop Model Updating

Closed-loop (reservoir or asset) model updating integrates the principles of
(optimal) control theory and closed-loop optimization with (multiple) data
assimilation into a workflow for reservoir optimization in terms of recovery
or financial measures over the life of the reservoir using periodic, near real-
time updates. A closed-loop controller operates with a so-called negative
feedback loop that dynamically compares the system output with the refer-
ence point using sensor systems. The measured difference is channeled into a
control device that dynamically applies the change to adjust the system input
so it better matches its output.
In oil and gas exploration, the concepts of closed-loop model updating in
a variety of workflows for reservoir management are presented in the open
literature under different names, however, with quite similar objectives as
stated above. Wang et al. (2007), Jansen et al. (2009), Chen et al. (2012),
Barros et al. (2015), and Sampaio Pinto et al. (2015) introduce closed-loop res-
ervoir management (CLoReM) (Fig. 6.7). Saputelli et al. (2006) and Sarma et al.
(2005, 2006) refer to the similar principles of optimal control and model
updating as real-time reservoir management. Dilib and Jackson (2013) describe
it as a closed-loop production optimization. Oberwinkler and Stundner (2005)
and Bieker et al. (2006) refer to the workflow as real-time production optimi-
zation, Saputelli et al. (2003) refer to it as self-learning reservoir management,
while Hanea et al. (2015) present the development and implementation
of the fast model update (FMU) workflow.
Frequently, the closed-loop model updating workflows as outlined
above integrate the principles of performing direct modifications of the static
geomodel parameters by the dynamic simulation modeling to perform the
geologically consistent model update and optimization using the computer-