## Abstract

Intelligent (a.k.a. smart or ‘flow control completion’) well (i-well) technology is being successfully used in thousands of wells worldwide. Dozens of companies are involved in manufacturing, modelling, installing, and optimizing i-well equipment. I-wells have their completion intervals divided into zones with packers, followed by regulation of production or injection in each zone using flow control devices. In production wells this balances flow profile as well as reduces inflow of unwanted fluids, overall resulting in increased oil or gas recovery. Finding the optimal locations of any given number of packers that improve the long-term production performance is a complex problem with few solution attempts published to-date. This paper is the first to offer an easy-to-implement, comprehensive solution to this packer placement optimization problem.

The complexity of this problem comes from the fact that ideally the i-well's designed flow control performance should be optimized together with its packer placement configuration, which makes it a largely multi-variable, mixed-integer optimization problem often requiring from thousands to millions of iterations to solve. Furthermore, where there is a long-term optimization objective, e.g. ultimate asset value or oil recovery, each evaluation of an i-well design requires a separate reservoir model run that takes time. This easily makes the i-well optimization problem insurmountable. The solution approach developed in this paper rests on two unique findings: the first one shows that optimality of a given i-well flow control performance reduces its subsequent, packer placement optimization problem to a simpler, perturbation problem which is much easier to solve. The second finding is a novel method relating an arbitrary, i-well flow performance to a mathematically equivalent non-linear electric circuit performance. This allows the well performance to be fast-solved semi-analytically using methodology from the circuit design theory.

Firstly in this paper, reduction of the i-well optimization problem to the packer placement -caused perturbation one is carried out by evaluating the impact of the annular flow isolation disturbance on the well/reservoir flow state. The relatively small weight of such disturbance on the scale of a simulation time-step is explained and further demonstrated on a test model. Secondly, an i-well flow model is presented as a nodal system, equivalent to a non-linear electric circuit, and subsequently solved using an approach adopted from the circuit theory. This solution is verified by numerical modelling. Lastly, these two findings enabled development of a fast, packer placement optimization algorithm that is described, verified, and finally successfully applied to a basic, test model.

The packer placement optimization algorithm presented in this paper, as well as the novel i-well nodal analysis approach, complement and complete the multiple i-well performance optimization workflows published to-date, all of which use the simplifying ‘infinite number of packers’ assumption for the lack of anything better. This algorithm will be of immediate use to the well and reservoir engineers involved in design, installation, and evaluation of i-wells.

The complexity of this problem comes from the fact that ideally the i-well's designed flow control performance should be optimized together with its packer placement configuration, which makes it a largely multi-variable, mixed-integer optimization problem often requiring from thousands to millions of iterations to solve. Furthermore, where there is a long-term optimization objective, e.g. ultimate asset value or oil recovery, each evaluation of an i-well design requires a separate reservoir model run that takes time. This easily makes the i-well optimization problem insurmountable. The solution approach developed in this paper rests on two unique findings: the first one shows that optimality of a given i-well flow control performance reduces its subsequent, packer placement optimization problem to a simpler, perturbation problem which is much easier to solve. The second finding is a novel method relating an arbitrary, i-well flow performance to a mathematically equivalent non-linear electric circuit performance. This allows the well performance to be fast-solved semi-analytically using methodology from the circuit design theory.

Firstly in this paper, reduction of the i-well optimization problem to the packer placement -caused perturbation one is carried out by evaluating the impact of the annular flow isolation disturbance on the well/reservoir flow state. The relatively small weight of such disturbance on the scale of a simulation time-step is explained and further demonstrated on a test model. Secondly, an i-well flow model is presented as a nodal system, equivalent to a non-linear electric circuit, and subsequently solved using an approach adopted from the circuit theory. This solution is verified by numerical modelling. Lastly, these two findings enabled development of a fast, packer placement optimization algorithm that is described, verified, and finally successfully applied to a basic, test model.

The packer placement optimization algorithm presented in this paper, as well as the novel i-well nodal analysis approach, complement and complete the multiple i-well performance optimization workflows published to-date, all of which use the simplifying ‘infinite number of packers’ assumption for the lack of anything better. This algorithm will be of immediate use to the well and reservoir engineers involved in design, installation, and evaluation of i-wells.

Original language | English |
---|---|

Article number | 108933 |

Journal | Journal of Petroleum Science and Engineering |

Volume | 205 |

Early online date | 15 May 2021 |

DOIs | |

Publication status | E-pub ahead of print - 15 May 2021 |

## Keywords

- Annular flow isolation
- Flow control device
- Intelligent or smart or advanced well completion
- Packer placement
- Wellbore flow modelling

## ASJC Scopus subject areas

- Fuel Technology
- Geotechnical Engineering and Engineering Geology