Aachener Beiträge zur Technischen Thermodynamik – serie

Ever since humans have existed, they have impacted the earth in many different ways (Redman, 1999). Currently, important impacts are associated with the excessive use of non-renewable fossil fuels such as coal, oil and natural gas. Most fossil fuels are used for electricity generation, heating and mobility (eia, 2011), and as feedstock in the chemical industry (IEA et al., 2013). Moreover, the use of fossil fuels is associated with carbon dioxide emissions (CO2) (IEA, 2014; Leimk¨uhler, 2010). Emitting CO2 into the atmosphere leads to global warming and disrupts the natural carbon cycle (Stocker et al., 2013). To close the disrupted carbon cycle, CO2 can be captured and re-utilized, thereby mitigating global warming and saving fossil resources (Styring et al., 2014).CO2 can be captured from current anthropogenic CO2 sources or directly from the atmosphere. Captured CO2 can then be utilized as valuable physical product “as such”or as alternative carbon feedstock for fuels, chemicals and materials. The general concept of CO2 Capture and Utilization (CCU) can be considered established: already today, CO2 is captured and utilized in processes in the chemical industry (Aresta et al., 2014). However, the scope of CO2 utilization is limited. Despite the existing industrial implementations as well as continuous progress and current efforts in CCU research, most CCU technologies are still in early stages of development. Besides the limited technological readiness, CCU is intrinsically challenging since both capture and utilization of CCU typically require substantial amounts of energy (Sakakura et al., 2007). If the provision of energy relies on fossil resources, indirect CO2 emissions are caused. Therefore, the intuitively expected environmental benefits from using CO2 are not given by default (Peters et al., 2011b). In fact, it cannot be ruled out that a tediously accomplished CCU process is finally environmentally less sustainable than a conventional fossil-based route. Therefore, it is desirable to know whether a specific CCU process is environmentally favorable. For this purpose, a reliable environmental assessment of CCU is required.As indicators for the environmental performance of CCU, a large variety of approaches are proposed ranging from qualitative design principles (Anastas andWarner, 1998) and metrics for ‘green’ chemistry (Constable et al., 2002) to CCU-specific ad-hoc criteria (Peters et al., 2011b; M¨uller and Arlt, 2014). These approaches are rather intended to guide the development towards ‘sustainable’ CCU processes than to systematically quantify the actual environmental impacts. In contrast to these approaches, Life-Cycle Assessment (LCA) is a systematic and standardized methodology to analyze the actual environmental impacts of products and processes (ISO 14040, 2009). Although LCA is frequently advocated for the environmental assessment of CCU (Aresta and Dibenedetto, 2007b; Peters et al., 2011b; Quadrelli et al., 2011), it is not yet standard practice (Sch¨affner et al., 2014). The reasons for this are the complexity of LCA as well as the limited data availability of many CCU processes at early design stages (Quadrelli et al., 2011). In this context, this thesis pursues two major goals: First, the thesis enables and supports the reliable environmental assessment for CCU processes using LCA. To overcome the complexity of LCA and to enable LCA novices to apply LCA to CCU, a jargon-free introduction is presented for LCA in the context of CCU. Furthermore, a framework for LCA of CCU is derived to avoid severe pitfalls in LCA of CCU. A case study for CO2-based polymers illustrates the application of LCA as well as the size and origin of environmental benefits of CCU. The second goal of this thesis is to provide an LCA-based approach to support the design of environmentally beneficial CCU processes at early stages. In summary, the thesis is intended to facilitate the utilization of LCA for CCU from early design stages to industrial implementation.

Knowledge of the phase behavior of liquid–liquid systems is indispensable for the design and optimization of process equipment. However, the ab initio prediction of liquid–liquid equilibria (LLE) is not yet reliable. Therefore, experimental generation of LLE data is still necessary. Current methods for the characterization of LLE consume large amounts of material, and the large volumes lead to large diffusion paths and hence long equilibration times, unless the mixture is thoroughly stirred. The components are typically quantified chromatographically, because chromatographic methods allow for the analysis of a broad range of chemicals. Automated setups are able to reduce the overall system volume, but merely determine partition ratios. They often use spectrometric quantification, because only one component is of interest. The use of spectrometry instead of chromatography leads to shorter times for analysis, but restricts the range of chemicals to optically active compounds.In this work, the benefits of the automated setups and of the conventional characterization of LLE have been combined: a platform has been set up to reduce the need for human interaction from sample preparation, to chromatographic analysis, up to data processing and reduction. The system volumes could be reduced from typically 25mL in jacketed equilibrium cells to 1 mL. The developed calibration technique allows for a substantial reduction in material, because neither solvents nor standards are needed. Instead of long equilibration times, now analytics is the bottleneck in the characterization of LLE, because chromatographic methods are rather slow. Therefore, the use of model-based optimal experimental design was investigated for the characterization of LLE, so that the characterization time could be reduced by reducing the number of LLE experiments. The optimal designs concentrate on few distinct experimental settings, and a large fraction of the experimental effort is put on areas where the curvature of the binodal curve is large. By applying model-based optimal experimental design, the prediction accuracy could at least be improved by a factor 2. Consequentially, the number of experiments could be reduced by a factor 2 without losing prediction accuracy compared to conventional designs. The developed automated setup was used to investigate the phase behavior of novel biofuel-blends with water. It was found that the water solubility in pure biofuels was up to 4 orders of magnitude larger than in conventional fuels. However, by adding blend components, the water solubility could be drastically reduced. The biofuel solubility in water is much larger than the solubility of conventional fuels in water. The blend components did not influence the solubility of conventional fuels in water.

This thesis presents systematic extensions of the continuous-molecular targeting framework for the integrated design of working fluids and ORC processes. In the first step of the proposed method, a simultaneous optimization of working fluid and process allows for the identification of an optimal hypothetical working fluid. The second step of the method is used to search for real working fluids that lead to similar performance.

The design and optimization of complex chemical processes is a key challenge in chemical engineering and requires knowledge of the underlying kinetic model. This information can be obtained from experiments by inverting the reaction mechanism, which needs to be known therefore. Solving this inverse problem, however, is mathematically challenging, if not impossible, and the reaction mechanism is mostly unknown for novel compounds. Both of these challenges are addressed in the present thesis by proposing a novel methodology for forward reaction model development, which is based on exploring chemical space without the need for prior knowledge.The presently proposed methodology makes use of reactive molecular dynamics simulations to explore the chemistry of gas-phase compounds. In these dynamic simulations the chemical systems are allowed to evolve naturally, based on the atomistic interactions. During this evolution, bond formation and cleavage are traced based on the atomic connectivities and used to detect reaction events. For each reaction, molecular structures are extracted and high-level quantum mechanical calculations are used to predict reliable thermochemistry and kinetics. This novel chemistry exploration scheme is used to generate an ab initio reaction model for the well-studied high-temperature methane oxidation, which is used as a reference. The ab initio reaction model and a novel reaction pathway observed during simulation are validated against this reference and against high-level quantum mechanics.The comparison of the present ab initio reaction model obtained for high temperature methane oxidation to well-established literature reaction model shows striking agreement. This validation case demonstrates the potential of forward reaction model development using the present purely predictive methodology. Moreover, a reaction pathway previously not considered in kinetic modeling is discovered using the present chemistry exploration scheme and successfully validated in a detailed kinetic study.Potential extensions to the presented chemistry exploration scheme are derived, discussed, and implementations are outline. These extensions focus on the inclusion of effects resulting from microscopic balancing: Pressuredependence and reactions involving non-thermal intermediates. Conversion of high-pressure limit reaction models to pressure-dependent models is intended to be described by microcanonical properties obtained via transformation of canonical properties. A similar transformation is used to obtain information about hot reactions, i.e. the kinetics of non-thermal intermediates. Ultimately, these extensions will be implemented in the presently proposed chemistry exploration scheme to obtain even more accurate ab initio reaction models.In conclusion, the present thesis addresses the increasing need for reaction model development of novel chemical compounds by proposing a novel chemistry exploration scheme. The agreement of reaction pathways and rate constants with literature data reveals the potential of trajectory-based chemistry exploration for developing quantitative reaction models.

Today humanity faces a challenge: Global warming that is caused by emissions of CO2. To reduce CO2 emissions, usage of CO2 as a renewable carbon source has gained attention. In particular, the CO2-based synthesis of formic acid has been intensively studied, since in addition to utilizingCO2, formic acid could also act as chemical storage of fluctuating renewable energy. Developing such storage is paramount for any renewable energy system due to the fluctuating occurrence of e.g. wind. However, for the CO2-based formic acid synthesis, only a few processes have been developed. Thus, novel formic acid synthesis processes are developed in this thesis. The development considers process structure and solvents, as well as formic acid derivatives due to the unfavorable equilibrium of the CO2-based formic acid synthesis.Firstly, a process is developed for the formic acid synthesis with the widely investigated solvent dimethyl sulfoxide (DMSO). The work identifies that DMSO has an azeotrope with formic acid. To avoid the high energy demand that usually occurs in processes containing azeotropes, a novel process is developed that uses a co-solvent. Process simulations show that this novel process has an exergy demand on par with the existing state of the art CO2-based formic acid synthesis process.Secondly, a process is developed where formic acid or a derivative is used to store fluctuating renewable energy before reforming to carbon monoxide. Here, more than 100,000 combinations of formic acid derivative, solvent and process are considered by a novel design method. For the developed process, process simulations show that the exergy demand is significantly lower than for the literature benchmark.Thirdly, a novel process is developed for the synthesis of the formic acid derivative methyl formate. In this process, the supply of CO2 is integrated with the methyl formate synthesis. Usually, the CO2 supply and the synthesis are considered separate. By integrating these two steps, the energy demand and minimum selling price of methyl formate can be reduced.Finally, the thesis shows that the highest reductions in CO2 emissions and in minimum selling price are achieved with the co-solvent and the integrated methyl formate processes. Since both these processes are developed in this work, this thesis contributes towards utilization of CO2 as a raw material in chemical industries.

Optimal performance of chemical processes requires both optimized operating conditions and carefully selected molecules such as solvents. However, the search for optimal molecules and process concepts often has a limited focus: Either processes are optimized using a pre-defined set of molecules or molecules are selected for novel applications based on simplified process indicators. At the same time, the search for optimal molecules often relies on strongly simplified thermodynamic models that require experimentally determined group interaction parameters and confine the molecular design space. Overall, current design approaches often do not capture complex process trade-offs and are limited to prescriptive sets of molecules which likely results in suboptimal choices.To address the challenge of identifying optimal processes and molecules, this thesis presents an integrated computer-aided molecular and process design (CAMPD) approach. The design approach uses quantum mechanics (QM)-based property prediction by COSMO-RS and is thus independent of experimental determined group interaction parameters while not relying on group additivity. For reliable and fast evaluation of complex processes, advanced pinch-based process models are employed. These pinch-based process models account for the inherent trade-off in molecular properties while being both computationally efficient and accurate in comparison to rigorous process models. The integrated design approach in this thesis is stepwise extended from process-level molecular screenings towards molecular design for separation and reaction-separation processes. The application of the presented integrated design approach is illustrated for various examples of solvent selection and process optimization. In particular, process concepts and solvents are investigated for the purification of bio-based platform chemicals as well as the production of CO from CO2. Overall, this thesis successfully integrates COSMO-RS property prediction in CAMPD and thus significantly expands the range and applicability of current CAMPD approaches.

Global greenhouse gas emissions significantly accelerate climate change and adversely affect life on earth. To mitigate these effects, greenhouse gas emissions need to be cut-down by significant reductions in primary energy consumption. Primary energy consumption can be reduced by increasing the efficiency of energy supply, which is mainly fixed during the synthesis of energy systems. Optimal synthesis of energy systems can be realized by mathematical optimization, however the solvable problem complexity is limited. In contrast, energy systems encompass highly complex energy conversion technologies and an increasing number of time-varying operation conditions. Thus, the synthesis of real-world energy systems usually result in large-scale optimization problems, which are computationally prohibitive.In this thesis, a solution framework is proposed to enable large-scale synthesis of energy systems. The framework exploits the two-stage character of synthesis problems: the decision stage of optimal investment for technologies and the operation stage. Thus, the framework solves two consecutive subproblems with reduced complexity by applying time-series aggregation methods.Time-series aggregation is performed by systematic clustering methods. Using these clustering methods, the framework simultaneously identifies the appropriate period length, required number of aggregated periods and number of aggregated time steps per period for accurate synthesis. Based on aggregated synthesis problems, the proposed framework calculates feasible solutions with known accuracy. In addition, the framework measures the accuracy as error of the objective function and iteratively refines the aggregation until an accuracy criterion is satisfied.The proposed solution framework for synthesis of energy systems is applied to two case studies motivated by real-world applications from industry. All results indicate that few time steps and few typical periods are sufficient to identify feasible near-optimal solutions of large-scale synthesis problems. Moreover, results show that intuitive aggregation methods (e.g., monthly averages) are not beneficial - only sophisticated clustering methods allow a strong time-series aggregation with high accuracy for synthesis of energy systems. Overall, the proposed framework outperforms available solution software in both solution time and solution quality.

Adsorption energy systems can be driven by thermal energy from waste heat or the sun and thereby allow reducing fossil energy consumption and thus reduce global greenhouse gas emissions. Adsorption heat pumps and chillers can provide heating or cooling, adsorption thermal energy storage allows storing thermal energy. However, adsorption energy systems suffer from high investment costs due to low performance.Performance of adsorption energy systems strongly depends on the equilibrium properties of the working pair as well as heat and mass transfer mechanisms of the adsorption material in the adsorption energy system (adsorbent configuration).Evaluating new working pairs and adsorbent configurations is rather challenging: While the working pair's equilibrium properties can be determined with standardized measurement equipment, heat and mass transfer mechanisms cannot easily be determined, since they strongly depend on the full-scale adsorption energy system. Construction and operation of full-scale experiments requires high effort. Besides, often only small amounts of an adsorbent configuration are available, which are insufficient for full-scale experiments.To resolve these drawbacks, this thesis provides and validates a comprehensive method to determine the performance of working pairs and adsorbent configurations in adsorption energy systems from simple small-scale experiments. As a representative class of adsorption energy systems, adsorption chillers are investigated in this thesis. A small-scale Large-Temperature-Jump experiment is combined with dynamic modeling of the transient heat and mass transfer processes. Additionally, the experiment is extended by an infrared camera. The additional temperature information allows to distinguish and to determine the time-resolved effective heat transfer coefficient and diffusion coefficient in the heat and mass transfer model.The heat transfer and diffusion coefficients are inserted into a full-scale adsorption chiller model to predict the performance. Exemplarily, a commercial available silica gel and the adsorbent class of metal-organic frameworks (MOFs) are evaluated for an adsorption chiller application.The method is validated with experimental data of a full-scale prototype adsorption chiller and shows high accuracy. Furthermore, the method allows optimizing the adsorption chiller for a given working pair or adsorbent configuration and allows identifying bottlenecks and potential for improvement of the working pairs. In summary, this thesis bridges the gap between small-scale experiments and modeling of full-scale adsorption energy systems. The method allows for a comprehensive and reliable evaluation of working pairs and adsorbent configurations for adsorption energy systems.

Synthesis of energy systems is a complex design task with a plethora of decision options. To evaluate these decision options, mathematical optimization is often used to identify the optimal solution. However, for decision support, more information than just the optimal solution is required. The decision maker needs to know design alternatives and their trade-offs to make a well-informed decision. Hence, mathematical optimization should be used as tool to generate multiple design alternatives. One way to generate design alternatives is the exploration of the near-optimal solution space.In this thesis, a decision support system is proposed for decision support by analysis of the near-optimal solution space. The near-optimal solution space consists of the near-optimal design space and the near-optimal objective space. For exploration of the objective space, a method is proposed to efficiently identify solutions which reveal trade-offs in the objective functions. For the design space, a method is proposed to span all near-optimal design alternatives by minimizing and maximizing design variables. The decision support system provides a holistic analysis of the near-optimal solution space by combining solutions from the near-optimal objective space and the design space. All generated solutions are analyzed to reveal feasible ranges of variables and objective functions. Additionally, the analysis determines trade-offs between decisions in both the design space and the objective space. Based on the results of the analysis, the decision maker can derive preferences. In an interactive feedback loop, these preferences are added to the synthesis problem to support the final synthesis decision.The proposed decision support system is applied to two real-world case studies. The first case study originates from pharmaceutical industry and focuses on the supply side of an energy system; the second case study is a retrofit of an urban energy system and also takes into account demand-side measures such as investments in insulation. For these two entirely different case studies, the decision support system provides decision support by identifying feasible designs, their costs and emissions, and the most important design trade-offs. Thereby, the decision maker is enabled to take well-informed decisions in the synthesis of energy systems.

The conductor-like screening model for realistic solvation (COSMO-RS) is applied to study three case studies of complex chemical systems. In the first chapter, cellulose solubility in ionic liquid (IL) solvents is considered and a molecular model for cellulose is presented for the calculation of its solubility in ILs and their mixtures with organic molecular solvents. The model is based on cellobiose units which were themselves obtained by a conformation search for a cellotetraose unit. The conformations that make up the model hence can take into account interactions of a dissolved cellulose repeating unit while making sure that the conformations are so that the repeating unit resembles a unit in the middle of a cellulose chain. It is shown that the model contains representations of all of intra-molecular hydrogen bonds as well as open hydrogen bonding sites for accepting inter-molecular ones witnessed experimentally. Relative cellulose solubilities in IL systems with cations [Amim], [Apyr], [C2OHmim], [C2OHpyr], [C2OC1mim], and [C4mim] and anions [Cl], [Br], [N(CN)2], [CH3CHOHCOO], [CH3COO], [HOCH2COO], [(C6H5)COO], [H2NCH2COO], [C2N3], [(C2H5)2PO3OH] and [HCOO] have been studied. In addition, cellulose solubility in mixtures of [C2mim][CH3COO] and 14 molecular solvents could be reproduced. Finally, a solvent screening over a large database of cation/anion pairs was carried out, providing for a design and discovery guide pertaining to novel IL systems for cellulose dissolution.

In order to slow down global warming, greenhouse gas emissions must be reduced. Human-caused greenhouse gas emissions come primarily from consumption of fossil energy. In order to reduce the consumption of fossil energy, the demand is rising for energy-efficient technologies. One promising energy-efficient technology is the adsorption dishwasher that was commercialized recently. The use of adsorbents enabled the adsorption dishwasher to save 25% of energy compared to a conventional dishwasher. To increase the savings and to further enhance the entire process, the adsorption dishwasher should be improved. The improvement should foremost focus on the adsorbents, since adsorbents are the key of this energy-efficient technology.This thesis therefore assesses adsorbents for the application in an adsorption dishwasher. The assessment is carried out both experimentally and theoretically. Theoretical investigations are divided in 3 stages of complexity:Stage 1 is a static analysis that is used to determine the required minimum mass of adsorbent. Stage 2 is a simple dynamic model that is used to determine the drying times. This simple 2-stage theoretical investigation method is applicable for any drying process in order to estimate the suitability of adsorbents. As a parameter study, adsorbents out of 3 material classes are evaluated regarding drying time and adsorbent mass required for the application in an adsorption dishwasher . The trade-off between drying time and adsorbent mass is discussed by employing the Pareto-frontier.Based on the results of the simple 2-stage theoretical investigations, suitable commercially available adsorbents are investigated experimentally. Evaluation criteria of the experimental investigations are working capacity, pressure drop over the adsorbent bed and dehumidification rate. Based on these 3 criteria, the most suitable commercially available adsorbents are identified.Finally, to assess adsorbents considering all dynamic interactions within the adsorption dishwasher, a theoretical investigation is conducted as Stage 3. Stage 3 is a complex dynamic model of the adsorption dishwasher including all its components. The tradeoff between drying time, adsorbent mass and energy consumption is discussed by employing the resulting Pareto-frontiers.In summary, this thesis presents methods for characterisation of desiccants. By using these methods, more suitable adsorbents are found for use in the dishwasher application.

In recent years, spectroscopy has developed into an increasingly valuable tool to determine the composition of mixtures; for scientific questions as well as for the industry. The increasing use of spectroscopy raises the question how to best use the obtained data. For the analysis of spectral data, the method of Indirect Hard Modeling (IHM) has been established besides statistical methods like PLS. IHM is a nonlinear method that can therefore efficiently treat nonlinear effects such as peak-shifts. In the present work, the IHM method is expanded to increase its applicability.IHM treats nonlinear effects in the spectral evaluation. Therefore, the direct proportionality between the concentration and the Raman signal of a component can be used for calibration. The resulting linear calibration model allows for reliable extrapolation. Thus, IHM can be used to study reactive systems, even if only binary subsystems can be used for calibration. However, thermodynamic systems with intermediates can so far only be calibrated by using thermodynamic models. In this work, a method is established that calibrates a reactive system with intermediates only based on the reaction mechanism as well as stoichiometry and electroneutrality.Spectral backgrounds, e.g., fluorescence, can be treated by a spectral pretreatment or via background models. However, spectral backgrounds are still a common source of error in IHM. Derivatives have long been used very effectively in statistical methods. Therefore, IHM is adapted so that it becomes possible to evaluate the first derivative of spectra.The calibration of IHM is mostly limited to the relative spectral intensities of the involved components. In the present work, a method is presented that uses the information in the calibration spectra more thoroughly. For this purpose, nonlinear effects are parametrized as a function of concentration.The commonly used peak profiles do not reflect the physical reality at a detector very well. As a result, narrow modelled peaks may change their apparent intensity if they are shifted. To correct these shortcomings, a new peak model is proposed in this work that is more closely aligned to the physical reality of a detector.

Consequential Life Cycle Assessment (CLCA) aims at capturing the environmental consequences of decisions such as the introduction of a new technology, the implementation of a policy, or the purchase of a product. CLCA combines technical and economic modeling approaches to track the consequences of decisions throughout the economy, considering both technical relationships within industrial production systems and market-mediated effects. However, although CLCA is well defined at a conceptual level, a commonly accepted modeling framework for CLCA is still missing, leading to wide differences in CLCA practice.To promote the systematization of the CLCA approach, this thesis proposes a comprehensive modeling framework for CLCA: the Technology Choice Model (TCM). Compared to existing approaches, TCM captures market-mediated effects in multiple markets at a substantially higher level of technical detail, while systematically considering constraints in factor availability, uncertainty, and suboptimal decisions. Due to its higher level of technical detail, TCM can model changes in technology mixes through both capacity adaptions and substitution effects among competing technologies. These changes in technology mixes are shown to substantially affect the CLCA results in two illustrative case studies on the introduction of new technologies and climate policy. Furthermore, the consideration of uncertainties and suboptimal decisions provides the basis for a first comprehensive uncertainty assessment in CLCA.The practical application of TCM is demonstrated in a large-scale industrial case study on novel Carbon Capture and Utilization (CCU) technologies in the chemical industry. These technologies use carbon dioxide from industrial point sources or ambient air as alternative carbon feedstock for chemical production. The case study shows that CCU in the chemical industry can reduce up to 3.5 Gt CO2-eq greenhouse gas emissions per year by 2030 and highlights potential barriers for CCU implementation. The results provide a strong scientific basis for the integration of CCU into international policy frameworks and research agendas. The application of TCM in this case study demonstrates the ability of CLCA to provide sound environmental decision support.

Diffusion and liquid-liquid equilibria (LLE) data are of major importance for the design of chemical processes as, e.g., extraction. Unfortunately, predictive equations for diffusion and LLE are not yet sufficiently accurate. The development of predictive equations suffers both from complex molecular interactions that are not fully understood and from a lack of experimental data that would be needed for validation. The reason for the lack of experimental data is that diffusion and LLE measurements are burdened with large experimental effort and high sample consumption.In this work, we address the need for improved diffusion predictions and experiments as well as for improved LLE experiments. For this purpose, a model-based approach is combined with computer experiments and the development of new measurement methods. The work was conducted in strong collaborations with TU Delft, FAU Erlangen-N¨urnberg, and RWTH Aachen University.The integration of the model-based approach into the development of measurement methods allows for the thorough evaluation of the quality of experimental results. Thereby, necessary improvements in model and experiment are identified and implemented. As a result, effective and consistent diffusion experiments for liquids and gases are developed. The new experiments allow for a significant reduction of measurement times and sample consumption by an order of magnitude.In addition, optimal experimental designs (OED) for diffusion and LLE experiments are investigated. OED identifies experiments that provide most information on the property of interest. Thereby, the number of experiments can be reduced without losing accuracy. We identify OEDs for diffusion and LLE experiments that can reduce the experimental effort by an order of magnitude. In addition, we identify optimal combinations of diffusion and LLE experiments for the economic design of extraction processes.Thus, the contribution of this thesis is two-fold: On the one hand, efficient experimental designs and setups are developed that allow for a significant reduction of experimental effort. On the other hand, the demonstration of the successful application of the model-based approach can serve as a motivation for an increased integration of models and experimental setups in future developments of experiments.

The global demand for cooling increases and poses a challenge for the stability of electrical grids and the emission targets. Thermally-driven adsorption chillers can meet this challenge by substituting electricity with thermal energy. However, although adsorption chillers are primarily driven by thermal energy, practical experience shows that the electricity savings often hardly compensate the electricity demand of the auxiliaries, such as pumps and fans. Thus, smart systems integration is crucial. Systems integration of adsorption chillers is challenging due to several features: (1) intrinsic dynamics; (2) close interaction between various system components; and (3) strong coupling of design, control, and operating conditions. These features lead to a variety of degrees of freedom, which make it difficult to use simple design and control rules for systems integration.To move towards optimal systems integration of adsorption chillers, this thesis provides appropriate models and methods. First of all, a model of an adsorption chiller is developed, which is accurate enough to consider quantities needed on system level, but also computationally sufficiently efficient for model-based optimizations. To demonstrate the model accuracy, the model is calibrated and validated for two types of adsorption chillers, a lab-scale one-bed adsorption chiller and a commercial two-bed adsorption chiller.Subsequently, the presented model is used for two areas of systems integration: the integration in hybrid systems and the integration within the auxiliaries. To exemplify the integration in a hybrid system, an adsorption chiller is integrated into a CO2 vapor compression cycle to efficiently provide refrigeration with natural refrigerants only. The exemplified integration identifies a significant potential for energy savings, which can be fully exploited by optimal design and control of the hybrid system. The demonstrated integration in hybrid systems is also an example for a conceptual integration, which quantifies the potential of a specific application.To integrate the adsorption chiller within the auxiliaries, an optimization-based method is presented for detailed sizing and control of all system components including auxiliaries. The presented method is demonstrated for a solar-thermally-driven adsorption chiller system and determines both, an energetically and an economically optimal design and control. In summary, the thesis provides a comprehensive framework for optimal systems integration of adsorption chillers: from model validation over conceptual integration to detailed design and control.

The accurate prediction of reaction kinetics data, e.g., reaction rate constants, for the microgel synthesis is highly desired because its usage in model-based design approaches promises the development of more specialized microgels and enables new applications. The complexity and the diversity of the microgel synthesis, however, often prevent experimental approaches but depict challenges for prediction approaches, as well. Therefore, this thesis analyzes a wide range of aspects relevant for the microgel synthesis individually and proposes prediction strategies for each.Here, the focus is on three main aspects: (i) The prediction of reaction kinetics in homogenous liquid reaction environments by combining high-level density functional theory and COSMO-RS, (ii) improving the property prediction of ionic species by applying the Cluster-Continuum approach and analyzing the uncertainty of the reference data, and (iii) enabling the kinetics prediction for inhomogeneous reaction environments with ReaxFF reactive molecular dynamics (MD) simulations.The results show that macroscopic properties of microgels, e.g., the crosslinker distribution, can be linked to the elementary reaction kinetics. In addition, the prediction of the solvation free energy of ionic solutes dissolved in neutral solvents is achieved with a deviation of just 2.0 kcal mol−1, which removes a main bottleneck for reaction kinetics prediction of such ionic systems. Furthermore, analyzing the statistical uncertainty of rare events in reactive MD simulations revealed that just a few reaction events are sufficient to obtain rate constants of sufficient quality. This is a key finding for the study of the microgel synthesis in reactive MD simulations because the required large system sizes and limited computational resources prevent the observation of more events practically. Also, force fields and their parametrizations should be evaluated regarding the eligibility for studying the microgel synthesis reactions based on the correct description of the vinyl group.In total, this thesis develops and presents a toolbox for reaction kinetics predictions for the microgel synthesis, which allows for accurate predictions over a wide range of different synthesis systems and enables further modeling and design approaches.

Results of life cycle assessment (LCA) studies are affected by uncertainties from various sources. These uncertainties decrease the reliability of LCA results. While uncertainty cannot be avoided, the impact of uncertainty can be quantified using uncertainty analysis. Still, despite recommendations to consider uncertainties, most LCA studies neglect uncertainty analysis.This thesis aims to facilitate uncertainty analysis in LCA through practical applications and enhanced methods. In a first step, basic uncertainty analysis includes parameter variations and scenario analysis. Basic uncertainty analysis is feasible even for low data availability and can easily be integrated in every LCA study. More sophisticated quantitative uncertainty analysis can employ analytical uncertainty analysis based on Taylor series expansion. However, the current first-order Taylor series expansion gives only a linear approximation of the uncertainty and limits the validity of the analytical approach to small uncertainties. To overcome this limitation, analytical uncertainty analysis is extended towards a second-order approximation. The developed second-order approximation provides higher accuracy at slightly increased computational cost as long as the LCA problem is small-scale.To assess uncertainty in large-scale LCA, uncertainty information of the background system can be gathered from LCA databases. In the widely used ecoinvent LCA database, the deterministic default values are represented by the median values of the log-normally distributed uncertain data. However, LCA results from studies with and without uncertainty analysis differ if the input uncertainties are represented by the median values. Using mean values for any uncertain data instead of median values is proposed to ensure consistency and comparability across LCA results with and without uncertainty analysis.The presented methods for uncertainty analysis in LCA are applied to a case study concerned with a novel hydrogen production process by methane pyrolysis. The case study shows that the developed methods for uncertainty analysis can be applied easily and increase trust into the results of the case study. Therefore, the presented methods meet the goal of this thesis and will hopefully enhance reliability of LCA results.

To avoid greenhouse gas (GHG) emissions and mitigate climate change, low-carbon technologies must be used to provide renewable energy and replace fossil fuels. However, this system transition is very material-intensive and leads to high demand for critical materials. Copper is such a material that is essential for electrical applications and many low-carbon technologies. The production of copper itself is an energyintensive process. Thus, two challenges arise that are addressed in this thesis: the flexible process operation in a fluctuating renewable energy system and the avoidance of process-based GHG emissions.The flexible operation of electricity-intensive processes can support the power grid and provide economic benefits. Demand response (DR) describes operational adjustments based on an economic incentive, such as fluctuating electricity prices. Our initial analysis shows a large DR potential of two electricity-intensive process steps in copper production. To consider the DR potential of the entire production process and to capture the dependencies of the many process steps, we formulate a detailed scheduling model of a representative copper production process. The developed mixed-integer linear program (MILP) allows minimizing the electricity costs without reducing the production volume. This process-wide scheduling enables significant DR potential, reducing annual electricity costs by up to 14.2% and shifting large parts of the electricity demand.Avoiding process-based GHG emissions is challenging because fossil fuels are hard to substitute in some processes. These processes use fossil fuels as high-temperature process heat and as chemical reducing agents. A promising alternative for these use cases is hydrogen (H2), when H2 is produced from renewable electricity using water electrolysis (Power-to-H2). The oxygen produced as a by-product offers further benefits as it can be utilized in copper production. To optimally design a power-to-H2 system, we formulate a MILP that minimizes the total annualized cost. The resulting CO2 abatement costs are 201EUR/t CO2-eq, which exceeds the current prices of EU allowances. However, a sensitivity analysis shows great potential through further development of water electrolysis.Decarbonization through Power-to-H2 offers additional DR potential. Our scheduling model of the decarbonized copper production shows that DR strongly contributes to low CO2 abatement costs. Consequently, this work identifies the potential of decarbonized copper production that provides a critical material for low-carbon technologies and supports the power grid through DR.

Due to the growing awareness of climate change, the chemical industry is increasingly considering not only economic but also ecological criteria in process development. Thus, environmental assessment methods are required that can be applied in early process development stages to support decision-making. An accepted, ISOnormed environmental assessment method is Life Cycle Assessment (LCA). However, LCA requires detailed information on mass and energy balances, which is usually not available in early process development. Furthermore, when assessing emerging technologies, future changes not only of the technology itself but also in the background system, e.g., the energy supply, have to be considered.In this thesis, the example of sector coupling of the steel and chemical industries is first used to investigate how changes in the background system can be considered in the assessment of emerging technologies and how these changes affect the assessment result. Afterwards, a framework for predictive LCA is presented that allows LCA to be integrated at early stages of process development. In contrast to existing approaches from literature using machine learning for component-specific predictions, this work combines machine learning regression models with methods of automated process design. Thereby, process-specific predictions of environmental impacts are enabled only based on the molecular structure of the desired product and the reaction equation. For this purpose, descriptors are determined purely predictively using quantum mechanics and statistical thermodynamics, as well as process shortcuts. In addition, an encoder-decoder neural network is used to increase the information density in the molecular descriptors. As regression models, an artificial neural network and a Gaussian process regression are trained on a consistent data set.The method is exemplarily integrated into computer-aided molecular and process design and used for the design of ecologically optimal solvents. In addition, the process-specific prediction is discussed for the example of CO2-based methanol. The results show that the integration of process-specific features into the LCA prediction increases the prediction accuracy and enables process-specific predictions. The presented method thus enables the integration of ecological criteria in the early development of chemical processes.

Reaction models are important for many engineering tasks, such as the optimization of internal combustion engines or production lines in the chemical industry. The automated reaction space exploration method ChemTraYzer attempts to alleviate the often time-consuming reaction model development process. ChemTraYzer uses reactive Molecular Dynamics (rMD) simulations to find the underlying reactions of the reaction process. Coupling the rMD output to quantum-mechanical reoptimizations of reactant, product and transition state (TS) geometries and to transition state theory, allows ChemTraYzer to produce kinetic and thermochemical data with the low uncertainties,which are required for many reaction model applications. In this thesis, I address two major issues of the ChemTraYzer method. First, I introduce two acceleration techniques for the rMD simulations, the pressure-accelerated Dynamics (pAD) and ChemTraYzer-Temperature-Accelerated Dynamics (ChemTraYzer-TAD) methods, to extend the applicability of rMD to reaction processes that occur on longer time scales than the nanosecond. The acceleration techniques are designed to work with minimal a-priori knowledge of the reaction systems, as they are meant for the exploration of unknown reaction space. With the pAD method, I describe a methodology to choose elevated simulation pressure and temperature to speed up the occurrence of reaction events without simulating a biased reaction network. ChemTraYzer-TAD is a progression of the pAD method. ChemTraYzer-TAD allows for higher simulation temperatures than pAD and allows for high boost factors of 108 as demonstrated in our ChemTraYzer-TAD case study. Both methods are applied to case studies of low-temperature ignition reaction processes. Second, I improve the performance of the automatic TS searches, which are a bottleneck in the ChemTraYzer methodology, by introducing a recovery method for failed TS searches. Our new recovery method improves the overall TS search success rate by up to 10 percentage points to a total of 48%. Finally, I demonstrate the practical usability of ChemTraYzer for automated reaction space exploration using case studies of chlorinated dibenzofurane formation and decomposition processes.