A MILP-based approach for the analysis of cooperative relationships in the transition of the energy-intensive industry

Abstract

The Dutch energy-intensive industry relies heavily on fossil-fuelled energy sources for production processes, being responsible for one third of CO2 emissions in the Netherlands. The road to CO2 neutrality by 2050 requires a massive transformation of the industrial sector: sustainable supply and (re-)use of industrial heat, CO2 capture and storage and improved efficiency and circularity are some of the routes that the transition of industry may follow. In the light of the European Green Deal, a rapid mobilization of investments is required in order to reap their benefits before 2050. In this context, new forms of cooperation within energy-intensive industrial clusters have the potential to accelerate the delivery of innovative solutions. Multi-actor and technical dependencies offer synergetic possibilities that make collaborative decision-making of great importance in a situation of profound transformation. The challenge is to understand how to systematically manage and govern such inter-firm collaborations to favour the transition of industrial clusters, while balancing actor profit and cluster welfare. Formal governance mechanisms such as bi- and multilateral contracts play a critical role in this respect, regulating the distribution of costs and benefits of investment projects, while consolidating the exchange of the resulting products to reduce the uncertainty of flows. As a result, the adoption of different forms of contracts needs to be carefully considered when evaluating investment options for decarbonization. Although the transition of the energy-intensive industry has been studied from different angles, current approaches fall short in analysing and understanding the relation between the technical (hardware) level and multi-actor (and institutional) context of an industrial complex. However, integrated methods and tools are needed to identify optimal transition paths, while enabling a dynamic evaluation of investments decisions under diverse multi-actor configurations. This thesis addresses the problem by means of the following main research question: “What are the implications of implementing contractual structures in optimization models to support optimal investment decisions in the decarbonization of industrial clusters?” Answering this question requires an exploratory research design combined with a modelling approach. The initial exploratory steps involve gaining insight in the field of contractual governance using elements of Transaction Costs Economics and Game Theory. A case study analysis of the industrial cluster located in the Port of Rotterdam is conducted alongside the literature study, analysing key investment projects for decarbonization and the underlying contractual network. As part of the hard-to-electrify (and therefore hard-to-abate) sectors, the Rotterdam cluster attempts to transition to a new energy system by switching from fossil fuels to electricity, low-carbon hydrogen and green hydrogen. Thus, a focus on hydrogen as fuel and for power generation is established with respect to investments’ evaluation. As a final step resulting from the exploratory approach, hypotheses on the impact of contracts on the transition of industrial clusters are formulated, together with the experimental setup required for hypotheses’ testing. The information gathered in the exploratory steps serves as input to develop a novel methodology to incorporate contractual structures in optimization models. The output of these research steps is translated into an integrated optimization model of the Rotterdam cluster using Linny-R, a software tool developed by Dr. P.W.G. Bots that applies Mixed Integer Linear Programming (MILP) optimization. Besides the quantification of results of hypothesis testing, the model provides a proof of concept of the developed methodology to include contractual structures in the analysis of investment decisions for the energy transition. In addition, the model provides measurable insights in the effects of contractual structures on the transformation of industrial clusters, specifying the cash flow of the total cluster and its distribution among individual actors as the main metric for the evaluation of the economic performance. Optimal investment curves and resulting CO2 emissions are finally analysed as determinants of the cash flows obtained from the model experiments. As a result, implications of the implementation of contracts in optimization models are observed and discussed. As part of the research outcomes it is found that, without an explicit implementation of contractual agreements, results obtained from MILP may be characterized by a distorted allocation of cash flows, in which sacrifices of some actors in the form of negative cash flows are compensated by higher profits of other cluster’s members. These kind of results are clearly misleading in the identification of an optimal transition path. Thus, the proposed methodology adds value to MILP-based approaches. The model results show that contracts do have an effect on the transition of the industrial sector. In fact, adopting different contractual structures shifts investment decisions overtime, affecting the timescale of investments and the associated CO2 reduction. Bilateral contracts seem to yield the highest benefits for the modelled system, both from an economic and environmental perspective, as contract terms are tailored to the specific needs of the contractual parties. This makes off-takers less constrained than with multilateral contracts, whose advantages can be fully observed in thick markets. Although the findings suggest that there may not be a universal, one-size-fits-all approach in establishing a set of contracts that best supports investment projects under all possible conditions, the proposed methodology is effective in providing direction for the next steps to make decarbonization projects an investable prospect in the energy-intensive industry. More precisely, barriers identified via the implementation of contracts specify areas of need for targeted policy action, addressing the creation of greater economic incentives for the integration of investment options. With specific respect to hydrogen applications, regulated revenue support in combination with policies recognising the higher social value of low-carbon/green hydrogen products may be relevant options to take into consideration. From a managerial perspective, the methodology allows to identify and address dependencies among investment options which may strengthen or jeopardize each other. Relevant examples for hydrogen integration are the need for increased CO2 savings and/or the need to complement green hydrogen with additional measures to capture refinery fuel gases to improve the business case of such projects. Other implications consist in the need to carefully rethink the relationship between producers and off-takers to gradually adapt it to market developments. As a general conclusion, implementing contractual structures in MILP enables a more accurate representation of the behaviour of rational actors, allowing for a dynamic analysis of the system’s behaviour. Not only the incentives arising for each contracting party are taken into account in selecting optimal transition paths, but also synergies and externalities arising from the combination of technical options are optimized. This generates more valuable insights from a multi-actor and system perspective. This research suggests several research avenues that can be pursued. First and foremost, further research should be conducted to provide understanding of possible consequences of combining diverse policy interventions with contractual structures. It is possible that contracts will play a greater role in determining the revenue generation of investment options when more policies that provide specific economic incentives for decarbonization are introduced. Moreover, besides incorporating a higher level of detail and improving input data, adjustments to the proposed model could be performed to simulate a shift in contractual structures that follows the (hydrogen) market developments. Moreover, additional model expansions could allow to examine the consequences of blue and green hydrogen integration on the incumbent grey hydrogen suppliers. Finally, building upon the results of this research, the performed study could be replicated for clusters in different locations. A cross-case analysis would then reveal whether location significantly influences the identification of optimal contractual structures to support the energy transition.