Reducing fuel consumption from dredge vessels is always one of the priorities of the ship builder and the dredge contractor. In addition there is an increased awareness worldwide regarding exhaust emissions. CO2, NOx and SOx currently are or will in future be regulated strictly by international legislation and local authorities. Looking at a dredge cycle of a Trailing Suction Hopper Dredger (TSHD), every stage is far from stationary and in fact the dynamic variations are very severe.. Its performance is strongly influenced by weather condition, hydrological condition, river/sea bed profiles, soil types and characteristics, discharge method and discharge pipeline configuration. Although, nowadays, sophisticated automation is employed on subsystems for optimizing the dredging process, the dredger is still mainly under control of operators. The knowledge, the skill, the attitude of the dredge operators determine the performance of a TSHD and make it even more dynamic. In order to reduce fuel consumption and exhaust emissions, the impact from said dynamics need to be taken into account. This research is therefore conducted in order to know, capture, understand and be able to predict the behaviour of the energy system of a TSHD under dynamic load. Comprehensive onboard measurements have been executed to collect real time data on the energy system behaviour and exhaust emissions. After post-processing, consisting of signal synchronization, correction of the NOx-sensor time lag, filtering, signal organization and unit conversion, the results are presented the very first time. It is observed that, at constant nominal engine speed transient loads push the fuel consumption, air consumption and NOx emission away from the stationary lines. However seen at a larger time scale (in the order of stages per dredging cycle), the effects from transient loads are neutralized. The most important conclusion is: that in terms of total fuel consumption and total exhaust emissions, a dynamic loading of the energy system is not resulting in a penalty. Nonlinear time domain simulation models of a TSHD energy system (for dredging and sailing) are built in Matlab/Simulink®. The main system components and their dynamics are included, which makes the scope of the simulation models wide enough to cover all required energy systems. By means of matching and validation, the precision of the simulation models is ensured on both component and system level. Most of the components are modelled based on first principle concepts. They provide the required level of detail for understanding the system behaviour and emissions. In addition, the simulation models are well structured, providing easy removal of non-needed components, addition of new components and improvement of existing components. Through onboard measurements and simulation models, the dynamics of the behaviour of energy systems and in particular the emissions of a TSHD was thoroughly investigated in the time domain. Using normalization and linearization, the response of the energy systems to external disturbances and control commands are also investigated in the frequency domain. The linear model requires only a limited number of normalized derivatives and time constants and they are relatively independent of physical dimensions of the components. In a block diagram as presented in the thesis it can be easily traced how the disturbances propagate through the energy system and the sensitivities of involved parameters can be judged. The linear model has several advantages when compared to the nonlinear model. In the first place a linear model is generic. Further the required normalized derivatives and time constants can often easily be estimated beforehand, also because their first principle origin is made explicit. So this makes it possible that the order of magnitude of the frequency bands of any system can be grasped, even before there is an actual design. Finally the dynamic response can be explored using classical control methods. This would be useful in deciding whether a control system is needed and to determine what kind of control strategy would be effective. In summary, the three approaches (onboard measurement, non-linear simulation model and linear model) presented in this thesis provide an exclusive database and practical tools to know, to capture, to understand and to be able to predict the behaviour of the energy system and emissions of a TSHD. By further development, such as: measurement from more vessels, extending the scope of the model, increasing the precision of the model and increasing the level of details of the model, these methods can eventually be used for optimizing the design of a TSHD and reducing operational cost (fuel and emissions).