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

TL;DR

The paper tackles integrated energy-system planning under dual-time-horizon uncertainty by formulating REORIENT, a multi-horizon stochastic MILP that jointly optimizes investment, retrofit, abandonment, and operation. It advances the state of the art with a centred-point stabilised adaptive Benders decomposition tailored for MILP, enabling efficient solution of very large problems and providing convergence guarantees. Key contributions include integrating retrofit and abandonment with investment, extending stabilised Benders to MILP with adaptive oracles, and demonstrating substantial cost reductions (e.g., 24% lower investment cost in the North Sea) in a large-scale European case. The work is practically impactful for policy and industry by informing retrofit versus abandonment decisions for offshore oil/gas infrastructure and hydrogen networks, while delivering a scalable computational framework for planners.

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.

Figures (10)

• Figure 1: Integrated investment, retrofit and abandonment planning. The grey dotted box includes some technologies that can be invested in. The offshore oil and gas platform can be retrofitted or abandoned. Otherwise, it can keep producing. The Offshore Energy Hubs (OEHs) zhang2022_OEH can generate electricity, and produce and store hydrogen.
• Figure 2: Illustration of MHSP with short-term and long-term uncertainty. (blue circles: strategic nodes, red squares: operational periods, $i:$ index of the strategic nodes)
• Figure 3: Illustration of adaptive oracles. There are three iterations in this illustration. The $x$-axis in the graph shows the value of the first element of $\hat{x}^O_{ij}$. In the first iteration, subproblem 2 is solved exactly at $(\hat{x}^O_{21}, q_2)$, and an exact cut is generated. An inexact but valid cut and upper bound are then generated by the adaptive oracles for subproblem 1 via cut sharing. In iterations 2 and 3, the exact cut is for subproblem 1, and valid inexact cuts are generated for subproblem 2.
• Figure 4: Illustration of centred point stabilisation. We assume $x$ has only one element and the inexact oracles provide inexact but valid cuts and upper bounds. The centred point is relative to the two inexact cuts and the target $T_3$.
• Figure 5: Dynamic level set management

Theorems & Definitions (4)

• Lemma 1
• proof
• Theorem 1
• proof