Hydrogen is considered to be a promising replacement of fossil fuel-based energy for the future energy supply. The possibilities to use hydrogen are extensive; hydrogen can provide high temperatures for industrial processes, produce electricity, heat buildings and be a fuel for the mobility sector without releasing carbon. In order to implement hydrogen transition in the current economy, a hydrogen infrastructure needs to be established. Prior research has pointed out that after certain alterations, the current natural gas infrastructure can transport hydrogen. The natural gas infrastructure in the Netherlands is extensive, and the capacity is big enough to satisfy the Dutch hydrogen need. Additionally, considerable costs can be saved if the natural gas network is used. The costs to adapt natural gas pipelines to transport hydrogen is ten times lower than the costs of constructing new hydrogen pipelines. The common approach for the design of the new hydrogen infrastructure is optimisation. However, prior research in regard to network evolution has indicated that infrastructure evolution is characterised by path-dependency, lock-ins and network effects. These factors are neutralised in the current optimisation methods. It is reasonable to presume that these factors of network evolution will also have an impact on the network transition that is based on an existing network. The first reason to assume this is that the development of the hydrogen network is estimated to take 30 years, in which other developments are likely to occur. Second, investments that are made, are locked in the new infrastructures, as expenses made cannot be spent again in a different manner. Furthermore, the investments made determine future options for investment. Research into the transition of one network, which is based on an existing network, remains uncharted. This thesis will evaluate the effect of different tactics and strategies on the transition of a network from fulfilling one purpose, distributing natural gas, to another, distributing hydrogen while taking path dependency into account. An agent-based model that applies rule-based behaviour is constructed to answer the main research question, which reads as follows: How do different transition strategies for the transition of a natural gas infrastructure to a (partial) hydrogen infrastructure perform over time? To create this agent-based network, a representation of a network was made with several components, so-called nodes and edges. Nodes are the entry or exit points of the network, the production sites of gas (either natural gas or hydrogen), heavy industry, energy generators or the points where the gas is converted from the transmission network to the distribution network. All nodes have a utility score based on the type of node, and the distance of the node to the closest hydrogen point in the network. This allows for the calculation of the utility score of a specific edge. Edges are the connection between the nodes and represent existing pipelines, potential new pipelines or temporary connections in de form of tanks. The effect of four tactical choices on the transition behaviour of the network is tested in regard to the costs of the transition, the volume hydrogen that is delivered to the network and the volume hydrogen that is exported. The following tactical choices are evaluated: -Prioritise the network transition on local optimisation criteria, -Including new pipe to be constructed in the excising graph, -Prioritising the export of both hydrogen and natural gas, -Allocate the available budget over time in different patterns. Based on the results of experimenting with the tactics, the following four strategies are formed: - Minimise cost, prioritise the export of both hydrogen and natural gas, - Minimise cost, no prioritisation of the export of hydrogen and natural gas, - Maximise hydrogen delivery, prioritise the export of both hydrogen and natural gas, - Maximise hydrogen delivery, no prioritisation of the export of hydrogen and natural gas. These strategies are applied to the random network developed for this thesis, and on topologies based on the Netherlands, Belgium and the United Kingdom. The results show that the strategies focusing on the minimisation of costs structurally have lower expenses than the strategies that maximise hydrogen delivery. However, in the case of the random starting topology and the topology based on Belgium, this is always at the expense of the hydrogen delivery as these strategies cause lock-ins. Prioritising the export of hydrogen and natural gas delays the developments of lock-ins and is therefore not only beneficial for the hydrogen export, but also for the volume of hydrogen delivered in the system. The topologies based on the Netherlands and the United Kingdom are less susceptible to lock-ins. There are situations in the topologies based on the Netherlands and the United Kingdom where the same volume of hydrogen is delivered in the strategies based on maximising hydrogen delivery. In these cases, minimising costs is the optimal strategy. In other situations, the hydrogen delivery in the strategies based on minimising costs is lower. In that case, a trade-off needs to be made between the hydrogen delivery and costs. \newline The experiments in this thesis have led to the seven insights that should be considered in the realisation of a hydrogen infrastructure. 1. The characteristics of a network are important. Best practices in one infrastructure should not be copied without any further consideration. 2. Purely adapting the excising network does not lead to the best outcome, and therefore the option for constructing new pipes on some critical points should be considered. The construction of new pipes helps to overcome lock-ins and therefore has a positive effect on the system outcome. 3. It is best to invest maximally according to the availed budget, the maximal capacity of the system and the foreseen future. With this, the system can benefit the longest from these investments and changes to the network. 4. Be reluctant about the network transition to certain geographic areas where the contribution is limited to only a small part of the network. 5. It is wise to determine minimal thresholds for the performance of the system to ensure that the system does not minimise costs at the expense of other key performance indicators. 6. Prioritise the flow of export and import of natural gas and hydrogen through the country. Not only does the country financially benefit from an export corridor, there are also positive effects for the network as this export corridor ensures an available hydrogen connection throughout the country. 7. Specific for the topology based on the Netherlands and the United Kingdom; there are situations where the strategy that minimises costs reaches the same hydrogen delivery as the strategy that maximises the hydrogen delivery. This reinforces the first insight. The specific situation and location of nodes should be reviewed in order to determine the optimal strategy. There are some limitations to the model created in this thesis. First, the local optimisation is done based on the utility of a pipe. This utility has a direct connection to the utility of the nodes it is connected to. Calculating the pipe utility as the added gain for the whole system would strengthen this model’s approach. Second, the average betweenness centrality and closeness centrality does not show a relation with the effectiveness of the tactics and strategies. This is probably because centrality measures are calculated for the whole system and not for the flows of hydrogen and/or natural gas. It is recommended to recalculate the two centrality measures, taking the gas flows into account, and observe whether there is a relation that can be used as a predictor for the effect of tactics and strategies. In this thesis, a system-level approach with a step for step transition is used. Network evolutionary elements liken path-dependency, lock-ins and network effects were taken into account. Including these elements of network transition, led to seven insights regarding the process of (network) evolution, compared to overall system optimisation. These seven insights should be considered when formulating an approach for the realisation of a hydrogen infrastructure.