[Paper Review] Integrated investment, retrofit and abandonment energy system planning with multi-timescale uncertainty using stabilised adaptive Benders decomposition

Hongyu Zhang, Ignacio E. Grossmann+4 more|Edinburgh Research Explorer (University of Edinburgh)|Mar 17, 2023

Global Energy and Sustainability Research4 citations

TL;DR

This paper proposes the REORIENT model, a multi-horizon stochastic mixed-integer linear program integrating investment, retrofit, and abandonment planning for energy systems under short- and long-term uncertainty. It employs an enhanced stabilised adaptive Benders decomposition to solve large-scale problems efficiently, reducing North Sea investment costs by 24% compared to conventional models and accelerating solution time by up to 6.8×.

ABSTRACT

We propose the REORIENT (REnewable resOuRce Investment for the ENergy Transition) model for energy systems planning with the following novelties: (1) integrating capacity expansion, retrofit and abandonment planning, and (2) using multi-horizon stochastic mixed-integer linear programming with multi-timescale uncertainty. We apply the model to the European energy system considering: (a) investment in new hydrogen infrastructures, (b) capacity expansion of the European power system, (c) retrofitting oil and gas infrastructures in the North Sea region for hydrogen production and distribution, and abandoning existing infrastructures, and (d) long-term uncertainty in oil and gas prices and short-term uncertainty in time series parameters. We utilise the structure of multi-horizon stochastic programming and propose a stabilised adaptive Benders decomposition to solve the model efficiently. We first conduct a sensitivity analysis on retrofitting costs of oil and gas infrastructures. We then compare the REORIENT model with a conventional investment planning model regarding costs and investment decisions. Finally, the computational performance of the algorithm is presented. The results show that: (1) when the retrofitting cost is below 20% of the cost of building new ones, retrofitting is economical for most of the existing pipelines, (2) platform clusters keep producing oil due to the massive profit, and the clusters are abandoned in the last investment stage, (3) compared with a traditional investment planning model, the REORIENT model yields 24% lower investment cost in the North Sea region, and (4) the enhanced Benders algorithm is up to 6.8 times faster than the level method stabilised adaptive Benders.

Motivation & Objective

• To develop an integrated planning framework that simultaneously addresses capacity expansion, infrastructure retrofitting, and decommissioning in energy systems.
• To model both short-term variability (e.g., time-series energy demand and generation) and long-term uncertainty (e.g., oil and gas prices) in energy system planning.
• To improve computational efficiency for large-scale stochastic mixed-integer programs arising in energy transition planning.
• To evaluate the economic viability of retrofitting existing oil and gas infrastructure for hydrogen use in the North Sea region.
• To compare the performance and cost outcomes of the proposed model against conventional investment-only planning models.

Proposed method

• Formulates a multi-horizon stochastic mixed-integer linear program (MIP) to represent investment, retrofit, and abandonment decisions across multiple time stages.
• Integrates short-term uncertainty via time-series parameters (e.g., wind and solar generation) and long-term uncertainty via stochastic oil and gas prices.
• Applies a stabilised adaptive Benders decomposition algorithm that uses adaptive oracles and level-set management to accelerate convergence.
• Employs a progressive hedging-inspired stabilization technique with a dynamic convergence tolerance and a line-search strategy to improve dual bound updates.
• Uses a master problem that optimizes investment and retrofit decisions, while subproblems evaluate system feasibility and optimality under scenario-specific conditions.
• Implements a scenario decomposition strategy with a CP (cutting plane) subproblem to refine the master problem iteratively using optimality and feasibility cuts.

Experimental results

Research questions

• RQ1 Under what conditions is retrofitting existing oil and gas infrastructure for hydrogen production economically viable compared to building new facilities?
• RQ2 How does integrating abandonment decisions into the planning process affect the overall system cost and investment trajectory in the North Sea energy system?
• RQ3 To what extent does the inclusion of long-term price uncertainty and short-term operational variability improve the robustness of the energy system investment plan?
• RQ4 How does the enhanced Benders decomposition algorithm compare in performance and convergence speed to standard Benders decomposition for this class of problems?
• RQ5 What are the implications of platform cluster behavior—continuing oil production or early abandonment—under different cost and price scenarios?

Key findings

• When retrofitting costs are below 20% of new-build costs, retrofitting is economically favorable for most existing pipelines in the North Sea.
• Oil-producing platform clusters remain in operation throughout most of the planning horizon due to high profitability and are only decommissioned in the final investment stage.
• The REORIENT model reduces total investment costs in the North Sea region by 24% compared to a conventional investment-only planning model.
• The enhanced Benders decomposition algorithm achieves a speedup of up to 6.8 times compared to a reference Benders implementation, significantly improving computational scalability.
• The integration of retrofit and abandonment decisions leads to a more cost-effective and operationally flexible energy transition pathway.
• The model demonstrates robust performance under multi-timescale uncertainty, with stable convergence and high-quality solutions across diverse scenario sets.